Introduction
The baou déné live in the former cotton belt of south-western Chad. The last generation of these ‘women who have a lot of things’ is approaching eighty, and their numbers are declining rapidly. The baou déné are a cohort of wealthy women farmers whose affluence can be traced to the colonial-era policy of mandatory cotton cultivation. These women played a vital role in the colonial economy, which was based almost exclusively on cotton, and their imprint on the region remains visible nearly a century after the policy was introduced. Men descended from the baou déné live in their villages and farm on land they inherited from these women, with animals descended from their original herds, even though the Sara subgroups of south-western Chad are normatively patrilineal. Some families even attribute their existence to the baou déné, because these women used their wealth to acquire wives for their descendants and to expand their families.
The daughters, granddaughters and more distant descendants of the baou déné are more likely to be small traders or market women than farmers. These younger generations of women are not ‘women who have a lot of things’. They are referred to as mosso, which means ‘to fall down’; in this context, mosso refers to the precariousness of their small businesses and the frequency with which they go broke or lose their capital. Operating on razor-thin margins, the mosso are unable to accumulate wealth like the women who came before them. They talk in self-deprecating ways about the difficulties they have ‘getting up’ after a fall, or amassing the resources they need to try their hand at business again. The term ‘mosso’ is used playfully to reference the repeated attempts of women to gain a foothold in the economy, but the term is understood to be derogatory and patronizing and it carries a sting.
Both of these categories of women and modes of existence are present – one fading and one gaining prominence – in the contemporary landscape of south-western Chad. This region has long been the economic engine of the country. It was referred to by the colonial administration as ‘le Tchad utile’, or ‘useful’ Chad, a moniker that reappeared at the turn of this century with ExxonMobil's plan to develop the region's oil reserves. Throughout the colonial period and most of the twentieth century, cotton was the most important source of revenue for the government and the south-west was the centre of industrial cotton production.Footnote 1 At the turn of this century, the economy transitioned from cotton to oil with the building of a major pipeline financed in part by the World Bank (Leonard Reference Leonard2016; Magrin Reference Magrin2001). With the launch of the pipeline project, Chad traded its dependency on cotton for a dependency on oil, but the region retained its identity as the motor of the national economy.
What is analytically interesting about the presence of the baou déné and the mosso as social categories is the near total absence of women as economic agents in the histories of the region or of these profit-making schemes. This paradox raises a number of questions. How can these groups of women be so present in the everyday life of the region while women as economic agents are so absent from the stories told about it? How can we understand the erasure of women as workers or economic actors from the records of these empire-building and profit-making projects? How is it possible that women in their productive capacities escaped the notice of colonial administrators and travel writers of the past and eluded the consultants, monitors and other chroniclers of a model twenty-first-century oil-as-development project? Why are the histories of ‘useful’ Chad silent when it comes to the economic lives of women, especially since, as the categories themselves suggest, these aspects of women's lives are so central to their identities? In this article, I draw on the insights and methodologies of feminist and critical historians, especially Michel-Rolph Trouillot, to explore the historical processes that produced these silences in the record and to show that these processes were different in the two eras and development projects.
I argue that, in the colonial era, colonial officers and other history-tellers were not able to imagine that women were fully engaged in cotton production, and especially that they were doing the back-breaking work of clearing tracts of forest and transforming them into cotton fields. This argument parallels, in some ways, an earlier one made by Judith van Allen about Igbo women's invisibility as political actors, which she saw as reflective of the tendency of British colonial officers, missionaries and scholars to imagine that ‘politics was a man's concern’ (van Allen Reference van Allen1972: 165).Footnote 2 Women's role in cotton cultivation was vital to the imperial project in Chad because mandatory cotton cultivation was implemented concurrently with the conscription of staggering numbers of men to serve as military recruits, porters and labourers for concessionary companies and on public works projects. This dual labour regime changed patterns of household labour, just as it did in other labour-exporting zones, where, in the absence of men, women were forced to take up roles that were considered to be men's work (Isaacman and Roberts Reference Isaacman and Roberts1995). The labour demands of growing cotton, especially the mandate to clear patches of virgin or long-fallow forest in every season, obligated some women to ‘change into men’, as it was expressed in the vernacular, but it also opened up opportunities for women to acquire land. That land later became a source of wealth and the basis for the emergence of several generational cohorts of women who ‘had a lot of things’.
In this century, the silences surrounding women's economic lives stem from a different set of dynamics. Global oil companies and their financiers were able to produce a narrative about women in the former cotton-growing zones that became hegemonic: this was that women in these zones farm just like men and that this feature of social life is embedded in the ‘culture’ of the region. The archival power of these claims was magnified by the modes of informational governance the oil companies and their financial backers adopted, specifically by the replication of these claims in different formats and their diffusion around the world as part of a commitment to transparency. While these claims might be read as a belated acknowledgement of the central roles women played in the cotton fields, they obscure the very different experiences of women in the era of oil. By the time the pipeline project got under way in 2000, the claims no longer reflected women's status as farmers. The cotton regime transformed the ecology of the region, and the pipeline project exacerbated land pressure by taking thousands of hectares out of production and introducing incentive structures that led to land hoarding. Some women had already been moved off the land by 2000, but the oil companies’ policies, which they advertised as gender neutral, had the effect of further marginalizing women and moving more of them off the land and into precarious trade.
The baou déné and the mosso are the products of these profit-making schemes. Here, I trace the historical processes that led to the emergence of these two groups of women – women who ‘have a lot of things’ and women who ‘fall down’. In examining the processes that made these categories of women with their distinct subjectivities and modes of existence possible, this article attempts to contribute to the writing of alternative histories of south-western Chad and to place women squarely within them.
Writing women into the past: a note on methods
Archives are problematic as sources of information about the past. Governance, efficiency and profitability all depend on information and on the ability to access, accumulate, record, manage, manipulate and disseminate it, whether the context is imperialism or corporate expansion (Mol Reference Mol2006). The archives of the colonial state and global oil companies conceal as much as they reveal, even – and maybe especially – in the context of regimes of transparency (Barry Reference Barry2013). They tell us as much about the people, states or companies that created them as they do about the people and processes they purport to be sources of information about. For this reason, rather than treating archives as neutral repositories of facts, feminist historians and other critical scholars have followed Ann Stoler's injunction to shift from ‘archive-as-source to archive-as-subject’, and have trained their attention on how sources are constructed, seeing those sources as sites not of knowledge retrieval but of knowledge production (Stoler Reference Stoler2002: 87).
To read the archive ethnographically means treating it as a cultural artefact with its own system of logic (Burton Reference Burton2003). Scholars have advocated the adoption of different reading practices that reveal these logics and the discipline archives impose on the reader. Stoler (Reference Stoler2009) advocates reading archives ‘along the grain’ as a way to imbibe the conventions, logics and rhythms by which colonial administrations operated. Others suggest reading ‘at an angle’ or between the lines to reveal subjects and areas that are hidden, obscured or silenced in a particular telling of history (Levine Reference Levine2004: 322). The point of these and other critical reading practices is to identify the various ways in which ‘power gets into the story’ and silences are created (Trouillot Reference Trouillot1995: 28). This happened in different ways across the two periods of history and in the two profit-making schemes I explore.
Narrations about women in the official records of colonial rule are sparse and dispersed. One way in which women appear, though, is in connection with trouble. Commanding officers in French Equatorial Africa (AEF) wrote about women as a source of food insecurity and drunkenness because cotton farming threatened food supplies and women used millet to make alcohol.Footnote 3 Women whose husbands were recruited to fight in the Second World War were described as dependent on state subsidies, as using trickery to obtain benefits to which they were not entitled, and as the source of prostitution in regional towns.Footnote 4 In yet other accounts, officers depicted women as brakes on social and economic advancement because they wasted money from the sale of cotton on ‘perles’ and ‘verroterie’ – what might be called ‘bling’ – instead of buying consumer goods that colonial agents associated with progress and modernization.Footnote 5 These references to women were one way in which administrators developed an understanding of the colonial project – and of themselves and others in relation to that project. The fact that it was women who were so often identified in connection with trouble is not incidental; as Joan Scott (Reference Scott and Scott1996) reminds us, naturalizing and generalizing differences between men and women are critical to building hierarchies, and gender was only one hierarchy used to legitimate colonial rule.
In the AEF, colonial authorities privileged the recording of certain kinds of information that also underscored difference across race and class. Difference and domination were exemplified not just in the existence of the archive – in who could write and record – but also through its stylistic conventions. Governors general responded to the reports of their lieutenants by complaining about missing elements, such as maps, correcting their grammar and spelling, and schooling them on the tone they should take and the importance of sending their reports to Brazzaville in a timely fashion.Footnote 6 However, their concern over the archive was about more than enforcing specific writing conventions: the archive was the embodiment of the colonial state, and the standardization of writing practices was a way of imposing order and projecting an image of mastery and control over a place that was sparsely occupied and, at least until mid-century, tenuously held. It was also a way of highlighting differences that underpinned the rationale for imperial governance in the first place (Aguirre Reference Aguirre2004).
Global oil companies have also developed distinctive ways of dealing with information, and these protocols were central to the pipeline project in Chad and reflective of the ‘information age’. The companies that operated in Chad produced an extensive archive of documents to demonstrate transparency and accountability, including via compliance with social and environmental standards established by the World Bank and other transnational organizations. These documents were produced in English and French and were readily available – by design – both in Chad and outside the country. Project archives included tens of thousands of pages of documentation in the form of risk assessments, plans for risk mitigation, monitoring reports and quarterly project updates that were curated for public consumption. The capacity of global oil companies and transnational financial institutions to produce and package certain kinds of information and disseminate it around the globe had the effect of amplifying some stories about women and their place in the local economy while burying others.
Critical historians have called into question the idea of an ‘official’ archive and the practice of privileging some sources of historical evidence, and this has given way to an expansion of the kinds of sources that go into the making of historical accounts. I looked for clues about women's economic lives during these two periods in multiple places. At the Archives d'Outre Mer in Aix-en-Provence and at the National Library in N'Djamena, I read the letters and reports of French colonial officers stationed in the AEF from the mid-1920s to the end of the colonial era and reports of special commissions set up to examine the prospects and problems of industrial cotton production in Central Africa. At the Centre d’Études et de Formation pour le Développement (CEFOD) in N'Djamena, I studied employee manuals, public relations brochures and other documents published by Cotonfran, the private cotton company that acquired monopoly rights to Chadian cotton. I also read colonial-era novels and travelogues.
Not much has been written about women's participation in the regime of mandatory cotton cultivation in Chad, or about women as farmers and economic agents in the era of oil, so these sources had to be complemented by oral histories. Over the course of multiple, extended periods of fieldwork conducted between 2000 and 2012, I collected dozens of oral histories from the last generation of the baou déné and their descendants and from women who are mosso.Footnote 7 Critics of oral history as history speak of memory as being less reliable than contemporaneously written reports or archives of the state; these critics might be particularly sceptical in this case given the advanced age of the remaining baou déné. Yet, I share the views of feminist historians such as Antoinette Burton who argue that women's memories are not merely supplements to the official archive but are ‘as legitimate, and as subject to interrogation, as official policy documents or other artifacts of the state, or, for that matter, other accounts of the past’ (Reference Burton2003: 26).
The history of the baou déné is a history in which, to use Trouillot's phrase, the silences have to be made to ‘speak for themselves’ (Reference Trouillot1995: 27). In narrating that history, I draw on multiple threads: ideas about separate spheres and the gendered division of labour that were taken for granted in the colonial era, reports about the conscription of male labour, census data, and oral histories of the baou déné and their descendants. In contrast, there was a strong but singular narrative about women's role in agricultural production in the time of oil. Narrating the history of the mosso begins, then, from the problem of an excess that stems from the demand for transparency but that has the effect of occluding other stories. In this contemporary case, I tell an alternative story by drawing on oral histories and fieldwork in the former cotton belt and on historical accounts of the disappearance of the forest that appeared more than half a century before work on the pipeline began.
Missing men and magical cotton
The emergence of the baou déné as a category is bound up with the history of mandatory cotton cultivation. A challenge facing the colonial administration in the AEF in the early decades of the twentieth century was how to make the colony – one of the poorest and most tenuously held in the French empire – self-sustaining. In Chad, the northernmost territory in the AEF, the answer was industrial cotton production. Southern Chad had sandy soils and a favourable climate for cotton growing – consistently high temperatures, annual rainfall in excess of 900 millimetres and a defined wet season. Despite the challenge of transporting goods from the interior of the continent to the coast, cotton could be grown profitably (although only under deeply exploitative conditions), and in 1928 the French launched a programme of industrial cotton production in the Moyen-Chari, the Logone and the Mayo-Kebbi. These three regions came to constitute Chad's cotton belt after attempts to expand production into other regions failed.
The second challenge the colonial administration faced, which was just as daunting as the first, was how to convince farmers to grow cotton. The response of farmers to mandatory cultivation was to flee or to passively resist. The transformation of Chad into a cotton-growing colony is a story of coercion and violence (Magnant Reference Magnant1986), just as it was elsewhere on the continent (Bassett Reference Bassett2001; Dampierre Reference Dampierre1960; Isaacman Reference Isaacman1996). Cotton was introduced under a system akin to what Sven Beckert calls ‘war capitalism’ – a mostly pre-industrial phase of mercantile capitalism based on imperial expansion, the expropriation of land and slave-like labour conditions (Reference Beckert2015: xv). The colonial administration made cotton cultivation mandatory and imposed an annual poll tax that could be paid only by growing cotton and selling it to La Compagnie Cotonnière Équatoriale Française, or Cotonfran. Cotonfran was a French and Belgian company created in 1926 and awarded monopoly concessions to the cotton produced in Oubangui and Chad.Footnote 8 Using methods pioneered in the Belgian Congo, Cotonfran worked with the colonial administration to distribute seeds and to purchase, process and export cotton; in Chad, however, Cotonfran relied on the government to use its authority to ensure that individual growers complied with the cotton regime (Cabot Reference Cabot, Coquery-Vidrovitch and Forest1983; Magnant Reference Magnant1986; Thompson and Adloff Reference Thompson and Adloff1960).
The introduction of mandatory cotton cultivation completely transformed the existing system of agricultural production in the cotton belt. In the precolonial period – and even up to 1928 – lineage groups carried out a form of shifting cultivation. These groups assigned adjacent plots of land to young, newly married men. Jean-Pierre Magnant referred to precolonial agriculture as ‘the work of men’ (Reference Magnant1986: 107) and described women as providing help in the labour-intensive planting season and in maintaining household gardens. According to Magnant, with the exception of widows, few women cultivated their own fields, and those who did used plots that had already been cleared to grow grains to make alcohol. When the soils became depleted, households and lineage groups moved to a new location, generally not far away, where they could open up virgin fields.
When cotton cultivation was initially introduced, payments were made to district chiefs, who were left to distribute them to individual farmers (Magnant Reference Magnant1986). Colonial authorities read farmers’ resistance to cotton not as a rejection of the mandate to grow it, but as a reaction to rapacious local chiefs who captured most of the benefits and left little or nothing for farmers. The administration's response included a combination of coercive measures and incentives. Government officials worked to fix people to the land, including through the use of reciprocal agreements with Cameroon to return people who fled across the border.Footnote 9 They also grouped farmers’ fields together so that growers could be more closely monitored and controlled and they put pressure on local chiefs to enforce the cotton regime (ibid.). But the most important reform the colonial government undertook to incentivize production was to individualize it.
Farmers were assigned to plots of land and paid directly for the cotton they produced. These units of land are known as cordes – named after the cords used to measure the plots allocated to each able-bodied adult, which varied in size from year to year but stabilized at roughly half a hectare (see also Dampierre Reference Dampierre1960; Magnant Reference Magnant1986).Footnote 10 Each year, farmers were assigned to clear new plots from virgin forest or long-fallow fields. This resulted in the de facto privatization of land in the cotton belt, as the act of clearing the forest and farming these plots made farmers – women as well as men – into people with rights in the land. Cordes designated for cotton cultivation were initially distinct from the lineage-based fields where food crops were grown; however, given the distance between these fields and the workload involved in growing two crops simultaneously, farmers began planting food crops on their abandoned cotton fields and over time the individualized mode of production introduced via the cotton regime replaced the lineage-based system (Magnant Reference Magnant1986).
Even as they allocated plots to individual family members, with one plot assigned to each adult, colonial administrators assumed a gendered division of labour in households and cotton fields. They imagined that men did the hard labour of felling trees, clearing land and preparing the plots for cultivation, while women and children helped with planting and harvesting. These assumptions about men's and women's contributions to cotton production were widely shared (see also Isaacman Reference Isaacman1996: 3). When Inspector Monguillot visited Chad in 1947 on a mission for the Inspector of the Colonies, he criticized Cotonfran for failing to supply Chadian farmers with basic tools in accordance with the terms of its monopoly contract. In support of his argument, he cited the recommendation from the Bangui conference on cotton from May 1944 that ‘each planter be given a hoe, of the type “Linda”, with a maximum weight of 400 grams, and that each man be also given either a machete or an axe’.Footnote 11 The Bangui recommendations encoded assumptions about the nature of men's and women's work, where men were presumed to need axes and machetes, while women needed only hoes.
Cotonfran also imagined a gendered division of labour, wherein women would be relegated to the domestic sphere and to helping their husbands in the fields. The company produced information brochures for prospective employees in which it described the ‘social centres’ it developed in the cotton belt where women could learn domestic arts, such as how to cook, sew, wash and iron as well as how to raise their children and basic principles of hygiene (Cotonfran 1958). Cotonfran and colonial administrators had ample evidence that women worked in the cotton fields, but they repeatedly downgraded their contributions or ignored information about the nature of the work they performed when it did not mesh with normative expectations about what women should be doing. In Chad, and throughout the AEF, fears about population decline and the supply of labour are difficult to disentangle from this gendered imaginary (Headrick Reference Headrick1994).
The effort to repress or reconfigure incongruous information about women and work was on full display in the report of Inspector Tupinier, who visited Chad in 1935, more than a decade before Inspector Monguillot arrived. The successive inspection teams were fixated on increasing yields, which had stagnated at 150 to 200 grams of cotton fibre per hectare. Finding solutions was tricky, however, because information about cotton production was patchy, including how much land was under cultivation, the size of the population in the cotton belt, and even who should count as a farmer. Take, for example, this excerpt from Inspector Tupinier's report:
The number of cotton farmers, as it was provided to the Inspector of the Colonies, has to be considered as providing only a ballpark figure. Certain heads of the subdivisions considered only able-bodied men as farmers. Others included women in the category of farmers, and even sometimes children. These are, without a doubt, those who understood correctly. Women collaborate with men in agricultural production. Sometimes this collaboration is the dominant form. Children, too, work alongside them, even if it is just for harvesting.Footnote 12
The inspector's questions might have been ambiguous to colonial officers, but the excerpt also suggests that field agents might have had different understandings or experiences of the nature of women's contributions to cotton farming. Confronted with a list of farmers that included the names of women, the inspector's response was to qualify and minimize the nature of their contributions. While it was possible for Inspector Tupinier to accept that women ‘collaborate’ with men or ‘work alongside’ them in the fields, just as children do, and even to acknowledge that ‘sometimes this collaboration is the dominant form’, it was not possible to think of women as cotton farmers on a par with men.
Nor was it possible at the time to imagine women as economic agents who accumulated wealth of their own. Jacques Rogué, the governor of Chad from 1943 to 1948, performed the exercise of calculating by how much yields would have to increase to justify mandatory cotton cultivation as a social and economic development project. In a letter to the governor general, Rogué presented the hypothetical case of a husband with two wives and three children who worked three cotton fields – one for each adult. Rogué calculated that if the family could produce 300 kilograms of cotton fibre per hectare – double the current yields – they would net 296 francs, and he produced a sample household budget to illustrate how this level of remuneration might cover the family's needs. What is notable is Rogué’s assumption that both household labour and income were pooled. The sample budget combined the income from the three fields and listed the expenses the husband – the presumed household head – would incur, including cloth, salt and hoes for his wives, and cigarettes ‘et divers’ (and other things) for himself.Footnote 13
A major problem with these assumptions was that the mandate for every adult to grow one corde of cotton was introduced at a time when another regime of forced labour was emptying the countryside of men. Thousands of men were being recruited by the colonial administration each year to serve in the military and the regional guard and to work as porters or labourers for private concessionary companies and on public works projects. This regime of forced labour predated the introduction of mandatory cotton cultivation but continued throughout most of the colonial period.
In 1925, Marcel de Coppet, then governor of Chad, described the region around Fort Archambault, the colonial capital of the Moyen-Chari, as devoid of men, especially during periods of labour recruitment. He described the town as ‘dead’ and said that he saw only women, children and old people on the roads.Footnote 14 The following year, the French novelist André Gide, a friend of Coppet, passed through Fort Archambault and noted that the town was once again tasked with sending 1,000 men to work on the Congo Ocean railroad, which Gide described as ‘a frightful consumer of human lives’ (Reference Gide1927: 200). Conscripts were held in guarded barracks and examined by a medical doctor who selected the most able-bodied among them. In 1927, the year before mandatory cotton cultivation was introduced, the commanding officer in the Moyen-Chari complained that his department was, once again, at the centre of ‘intensive’ recruitment efforts, and noted that more than 8,000 men had already been removed from the region.Footnote 15 While the Moyen-Chari was the most populous district in the colony, with an estimated population in 1927 of 58,000 men (sixteen years of age or older), the problems posed by persistent waves of recruitment were a recurrent theme in the reports of officers stationed there.Footnote 16
Colonial officers expressed concern about the demographic impacts of mass out-migration, the spread of sexually transmitted diseases, and high rates of mortality. In 1941, the commanding officer of the Moyen-Chari noted that, in the first half of the year, he supplied thousands of men for public works projects, including 2,100 for road projects and hundreds more for the army and to work as porters for private companies:
The effort of the Moyen-Chari, if we also take into account the hundreds of labourers and porters we are forced to provide for private companies, has been enormous, and we cannot continue to dip indefinitely into what is called this reservoir of men, because the army has also called for large numbers of men for many years now. It's all of the youth who are leaving for work.Footnote 17
Similar numbers of men were recruited in the second half of the year. The commanding officer noted that most of the previous recruits had not returned and he complained that ‘little by little Sara country is being emptied of its young people. What to write about all of these workers!’Footnote 18
Private companies were a major source of the demand for labour. During the second half of 1942, Sara men transported ‘tons and tons’ of merchandise from the region to distant cities.Footnote 19 The commanding officer in the Moyen-Chari noted that he had to devote considerable effort to recruiting porters for private transport companies and that in this six-month period alone a single subdivision had supplied 833 porters and was being asked for another 120. In a fit of pique he urged the governor to put an end to what he called the ‘moteur Sara’.Footnote 20 Other officers threatened not to respond to additional requests for porters or complained to the governor that it was impossible to satisfy the demand:
Also, during periods of intense agricultural labour, the transporter, the entrepreneur and the colonial official race either to the head of the subdivision or to the head of the department to demand that, right away, we supply 50, 60, 100 men, as though all we have to do is walk out on the pavement to give them satisfaction. Often these demands are delivered in a tone that we support only because we represent the administration.Footnote 21
The situation was similar in the Logone, where men also fled across the border to Cameroon to avoid recruitment. The report for the first semester of 1941 indicates that labourers in Garoua, a town in northern Cameroon, were almost all displaced Sara, and that military recruitment turned up 300 Sara out of 400 recruits.Footnote 22 Using language similar to his colleague in the Moyen-Chari, the commanding officer in the Logone described the region as a ‘pool of labour into which everyone comes to dip’ and complained that census-taking operations revealed no increase in the population because the men who left rarely returned to their villages or even to Chad.Footnote 23
The impact of conscription is indeed reflected in census data from the region. Counting people was one of the main tasks of colonial officers and census-taking was used as a way to mark the presence of the administration and ensure regular contact with villages. Reports to the governor general included tables noting the numbers of boys, girls, men, women, older people (‘vieillards’) and taxable subjects in villages visited by agents during that quarter. While population counts were known to be imprecise and subject to manipulation, they confirm the claims of colonial officers that men left or were recruited out of the region in large numbers. Census data show more boys than girls in most villages, although the gender gap among children is relatively modest. Among adults, however, women far outnumber men, even with older people separated out, and in some villages the male to female ratio approaches one to two.Footnote 24
Colonial officers often blamed the recruitment of men for stagnant production.Footnote 25 Yet they appear never to have asked how cotton production remained stable, even in hard-hit subdivisions, or increased, as it did in most places, despite the loss of thousands of able-bodied men every year. The government and Cotonfran noted that the rapid and regular increase in the production of cotton – from 5.2 million tonnes in the early 1930s to 97.9 million tonnes in the early 1960s – was ‘unique’ in francophone Africa (Présidence de la République du Tchad et al. n.d.: 28). How was such a remarkable expansion of cotton production possible alongside the massive recruitment of human labour that removed thousands of the most able workers from the region every year? Who did colonial administrators and Cotonfran employees imagine was doing the work of clearing new fields and producing more and more cotton if men were in such short supply?
The Ministry of Agriculture credited the expansion of production to extension staff and enhanced educational efforts targeting growers, while Cotonfran chalked it up to the conditions of its monopoly contract and the assurances it provided to farmers that the cotton they produced would be purchased (Présidence de la République du Tchad et al. n.d.). The company also attributed increased production to new seed varieties (see also Thompson and Adloff Reference Thompson and Adloff1960), although colonial officers complained that Cotonfran frequently distributed cotton seed late or provided poor-quality seed, and inspectors argued that the replacement of old varieties with new ones could only be accomplished gradually (Stevelinck Reference Stevelinck1953).
While some women worked the fields alongside their husbands, it is clear from oral histories, and even from contemporary living arrangements in the former cotton belt, that other women were forced to grow cotton to pay the poll tax. This was the case for Lelmbaidje, who was eighty-two when we interviewed her:
One part of my wealth is land. I have seven hectares. All of that land is the fruit of natural acquisition, with the help of axes and clearing. To grow cotton, women had to cut down trees and clear the roots from the field, which they did by burning the trees and then with axes. That is how most of them became landowners. I started paying taxes in 1956, or two years before the proclamation of the Republic of Chad.Footnote 26 After independence I continued to pay taxes … It was taxes that pushed most women of my generation to grow cotton, because it was the only way to have money at that time.
Many of the baou déné talked about the difficulty of clearing dense forest, with trees so tightly packed that they had to be burned until they could be cut with axes. For years, farmers struggled to make enough money from cotton even to pay their taxes.Footnote 27 In lean years, taxes could not be collected in full and officers in the subdivisions pleaded for relief or rate reductions. But women did not rely on cotton alone; the cotton regime opened up other opportunities for wealth creation for enterprising women. Because new plots of land were cleared every year, mandatory cotton cultivation extended the areas that were ‘fully prepared’ and available for growing food crops.Footnote 28 Women who acquired rights in the land by clearing these plots were able to become wealthy not just by selling cotton but by selling food crops and using the proceeds to invest in cattle or to hire labour to expand the areas they kept under cultivation. Anne, who was seventy-nine at the time of our interview, planted her first cotton field when she was seventeen and used a combination of strategies to accumulate wealth:
I decided when I was sixteen to have my own manioc field. The next year, I decided it would be better to plant cotton. With the money from selling the manioc I planted half a hectare of cotton. After selling it, I made, in 1954, a total of 19,000 francs. With that I bought a male ox for 7,000 francs and a female for 5,500 francs … After two years, in 1956, my cow gave birth to a pretty calf even though I was using her for work in the fields. That's how, after six years, I found myself with eight head of cattle. In 1962, I grew two hectares of cotton and made 396,000 francs. With that, I started my commerce and that's how I evolved. I became rich … At that time there was an abundance of land that wasn't used. Some women didn't have their eyes open to that opportunity to have things distinct from their husbands. But I cleared land … and some women imitated me too.
Some women who ‘had a lot of things’ supported their husbands and families and even lived in their natal villages with their husbands instead of adhering to the normative pattern of patrilocal residence.
The absence of historical sources about women as land clearers, cotton producers and independent economic agents may be explained in part by the fact that the colonial administration, which implemented the cotton regime for Cotonfran, was light on the ground in Chad. Graduates of the École Coloniale, France's finishing school for men planning to serve in the overseas empire, selected their post based on class and rank – and the AEF ‘was invariably considered the least desirable assignment’ (Cohen Reference Cohen1971: 46). The lack of personnel was particularly acute in Chad, where there was a ‘skeletal service’ (Debierre Reference Debierre1927) and entire subdivisions were ‘left to themselves’Footnote 29 or were administered by officers from neighbouring districts who were forced to do double duty.Footnote 30 This was reflected in their arm's-length reporting about how the policy of mandatory cotton cultivation was carried out.
A more compelling explanation, however, is that the idea that women cleared tracts of forest and turned them into fields was unimaginable in the colonial period. Colonial officers complained incessantly about recruitment, but despite constant references to the frightening numbers of men pressed into service and concurrent reports of increased cotton production, colonial officers, inspectors and travel writers never imagined women as independent farmers and economic actors. Trouillot notes that the creation of silences is not always conscious: ‘When reality does not coincide with deeply held beliefs, human beings tend to phrase interpretations that force reality within the scope of these beliefs. They devise formulas to repress the unthinkable and to bring it back within the realm of accepted discourse’ (Reference Trouillot1995: 72). Even when evidence surfaced of women farming – and not just collaborating or working alongside men – it was reinterpreted or sidelined. Inspector Tupinier's dismissal of women farmers on the lists provided by colonial officers is a case in point. In the same document, Tupinier described farmers waiting three days at the cotton market in Moundou to be paid, most of them without any food, because Cotonfran had run out of cash. Among the farmers he saw at the market were ‘breastfeeding women’.Footnote 31 In addition to being rare and fleeting, entries such as these about women as farmers invariably emphasized their domestic, maternal or dependent status.
The baou déné are not mentioned in the archives from the colonial period, nor is there any mention of women who amassed land and wealth. Even in contemporary writing on the cotton belt, women are rarely depicted as independent farmers. One notable exception to this rule is the writing of the American anthropologist Ellen Patterson Brown, who appears to have been the first chronicler of life in the cotton belt to have noticed the baou déné. She did so in the 1970s and 1980s – just as the conditions that made it possible for them to ‘have a lot of things’ were beginning to change.
What transparency masks
Brown wrote about the baou déné, though not by name, as part of a study of famine recovery published in the wake of the droughts and famines that swept Chad and the Sahel in the 1970s and 1980s. Her study of famine recovery was comparative, designed to look at how different ethnic groups coped with food shortages and how well their strategies worked. She concluded that the Sara recovered from the famines more quickly than groups in other parts of the country, in part because ‘men and women … hold similar rights and resources and labor at similar tasks in the field’ (Brown Reference Brown, Downs, Kerner and Reyna1991: 295). Brown wrote that most Sara women ‘adopt behavior which Sara culture defines as male – farming, providing money for household expenses, and acting as head of household (indeed a woman who acts this way for many years is called “tel dingam,” changed into a man)’ (ibid.: 299).
These claims took on new life when ExxonMobil hired Brown as a staff anthropologist at the turn of this century and decided to use her work to develop social and environmental risk mitigation policies for a major oil field development and pipeline project (Mallaby Reference Mallaby2004). The loss of land was expected to be the most pressing effect of this experimental project for farmers in the former cotton belt. The oil companies involved in the project proposed to mitigate the effects of land loss by paying farmers for lost labour, crops and trees – but not for the land itself, which they said belonged to the state. On the basis of Brown's research, the companies claimed that no special provisions needed to be made for women, since most women could and did farm just like men; those women who chose not to farm would be covered as dependants when their husbands lost their land (EEPCI 1999). Brown's research therefore led to the formulation of gender-neutral land acquisition policies, or policies that positioned men and women equally vis-à-vis compensation payments.
What Brown's research and the consortium's policies overlooked, however, was that the period from 1930 to 1980, the peak period of the cotton regime, was a remarkable interlude in the history of the rural south – one in which women could access land just like men. Women's ability to change into men was not a stable feature of social and cultural life but a response to a contingent set of conditions: mandatory cotton cultivation, the mass conscription of male labour, and an ample supply of arable land. What the consortium's policies neglected to take into account were the sweeping transformations in the rural landscape that had unfolded over the course of the twentieth century. Women had slowly but steadily been moved off the land and into more precarious occupations as soils were degraded by cotton cultivation, the forest disappeared, and ‘natural acquisition’ gave way to inherited land that passed through the patriline. In the latter decades of the twentieth century, more and more women began to engage in small trade to earn cash as a supplement to their earnings from agricultural production or in lieu of them. These women who relied on small trade came to be known as mosso, or women who ‘fall down’ – a reference to the precarity of their off-farm livelihoods.
This transition was partly foretold. From the outset of the programme of mandatory cotton cultivation, colonial officials worried about the speed with which forest clearing was taking place. As early as 1934, just a handful of years after mandatory cotton cultivation had been introduced, the administration noted that increased cotton production was only the result of extending the surface area under cultivation and urged that ‘from today, it is no longer increased production through the development of more fields that must be sought, but the improvement of yields per hectare’.Footnote 32 One year later, Inspector Tupinier, picking up on this theme, suggested that farmers should not immediately clear any additional land: ‘Colonial administrators were just about unanimous in their opinion that we were too quick to develop cotton fields and that the actual fields under cultivation should be considered for a certain time as the maximum.’Footnote 33 At the same time, colonial officials were under pressure to increase yields in order to make cotton even marginally profitable for farmers, who could not otherwise be motivated to grow it. The response was to assign farmers to grow cotton on the plots with the richest soils.Footnote 34 The annual clearing of new patches of virgin forest led to repeated warnings by colonial officials and inspection teams that the administration should anticipate the disappearance of the forest and be prepared to intensify production (Cabot Reference Cabot1955; Reference Cabot1965; Gilg Reference Gilg1970; Stevelinck Reference Stevelinck1953).
As the forest disappeared, farmers came to rely on inherited land. In her study of famine recovery, Brown minimized the importance of inherited land, writing that ‘inherited land is also of little importance; both men and women can and usually do clear new fields in virgin bush or in long-fallow fields’ (Reference Brown, Downs, Kerner and Reyna1991: 297). Yet, by 2000, when the pipeline project was getting under way, virgin bush and long-fallow fields had disappeared in most of the cotton belt (and in some parts of the cotton belt this had happened decades earlier). Land pressure was further exacerbated by the pipeline project in two ways. First, the project needed land on which to site infrastructure, and the consortium prioritized agricultural land to avoid displacing people from their homes or relocating entire villages. Second, the policies the consortium devised for taking land attached monetary value to labour, crops and trees and required farmers to stake exclusive claims to plots and to actively work them in order to be eligible for compensation payments and other benefits.
The process of divvying up land occurred within families and prioritized family members – disproportionately men – who were thought to be able to render the land legible to the oil companies as fields or as plots of land under cultivation (Leonard Reference Leonard2016). In tying compensation payments to labour, the project reduced the flexibility that had previously characterized land use arrangements in the cotton belt. Families hesitated over lending land to non-family members, since it meant risking the loss of a potential payment. Within families, men worried about the sudden wealth that might accrue to their wives if they granted them access to plots under their control (ibid.). Under these conditions, women found it increasingly difficult to access land. As land became not only scarce but valuable, claims to it increasingly referenced norms of patrilineal descent. Even the rights to land of the descendants of the baou déné – who were living in villages to which they were attached through their maternal relatives – were called into question.
These dynamics were not visible outside the oil field region, where the narrative that circulated about gender and land was much simpler. The consortium's gender-neutral plans for land expropriation were part of the Environmental Management Plan, a compilation of policies written in response to World Bank demands to reduce the impacts of the pipeline project on local communities. The Plan was produced in two languages and in multiple formats, some designed for local audiences but most for readers outside Chad. It was posted on websites, distributed through the World Bank's public information centres, handed out on CDs by consortium staff, and delivered locally to non-governmental organizations and libraries. Each time the consortium produced its quarterly project updates, it reiterated the terms of the Plan. Multiple teams of monitors and consultants tracked the consortium's adherence to the Plan and posted trip reports and updates online, repeating and recirculating claims about land use arrangements and about how women and men worked in the fields. The layering of monitoring and surveillance produced a voluminous amount of writing on the project, almost all of it authored by the project's architects, consultants and monitors, and all of it referencing and reinforcing the terms of the Plan.
Information, as Arthur Mol (Reference Mol2006) has noted, is a resource with the power to transform environmental debates. The oil companies operating in Chad produced policies and a system of oversight and surveillance to convince a sceptical public that the social and environmental risks associated with the project were minimal and could be mitigated successfully. The policies were designed to compensate farmers for their labour, regardless of who performed it, but they had the effect of concentrating land in men's hands. They also simplified and streamlined the process of land acquisition by ensuring that the oil companies did not have to sort out complex histories of social and gendered relations in the land. Transparency, as enacted through layers of monitoring and reporting, had the effect of amplifying and anchoring some stories of what was happening in the former cotton belt while obscuring or silencing others, including the story of the displacement of women from the land and the rise of the category of the mosso. Women in the time of oil, far from holding ‘similar rights and resources’ and labouring at ‘similar tasks in the field’ (Brown Reference Brown, Downs, Kerner and Reyna1991: 295), were moving off the land and falling down and trying to get back up.
Conclusion: gendered histories of precarity
Esther Boserup's landmark study Women's Role in Economic Development (Reference Boserup1970) presaged a burst of scholarly work on the gendered impacts of agricultural modernization. This work is virtually unanimous in suggesting that colonial policies marginalized women and undermined their economic possibilities by relegating them to subsistence or low-tech farming and domestic labour. The experience of women in Chad's cotton belt would seem to mark an exception to this general trend. The policy of mandatory cotton cultivation coupled with the conscription of thousands of able-bodied men to perform corvée labour left some women with little choice but to clear land and grow cotton – to change into men. But the cotton regime also transformed systems of land tenure in ways that provided openings for enterprising women to amass wealth independent of their husbands. Because these scenarios were unthinkable to colonial officers, there is virtually no mention of the critical roles women played in agricultural production and in the unprecedented growth of cotton production in Chad in colonial-era archives and therefore no way to understand the emergence of the category of women known as the baou déné.
In contrast, market-based development policies like those that underpin contemporary extractive industry projects are said to be liberatory for women because they individualize risks and rewards and, in doing so, erase gender-based distinctions. The policy in Chad was to pay farmers for their work on the land and the loss of crops and trees, regardless of sex. Land expropriation policies were based on assumptions about the ability of women to access land and to farm like men, but as the forest disappeared, so did the baou déné and the possibilities for women to amass land and wealth and to change into men. The precarity of the mosso stands in sharp contrast to the sustained wealth of older women.
The transition from the baou déné to the mosso can be read as a gendered history of precarity in the former cotton belt of southern Chad. This change appears to have gone unnoticed. It is absent from records and accounts of the region and of the major economic projects that have shaped it in the last century. Women's economic lives are absent from the archives for reasons that differ over the two periods and projects, but women were central to both and – in different but powerful ways – to the creation of profit, empire and ‘useful’ Chad.
Acknowledgements
I am grateful to Ngondoloum Salathiel for his help in conducting some of the interviews for this article.
