As the first project coordinator recounted, since the project’s preparatory missions the focus was on the possibilities for technical cooperation with the research institutes, rather than finding solutions to the cotton sector at large. However, even if the local research institutes were not formally submitted to the cotton companies, and neither were the latter the only existing conduit for technology transfer to farmers , they did become unavoidable. This was so both practically – the companies control the distribution of basic inputs – and politically, due to their muscle vis-à-vis both research institutes and governments, especially in those countries highly dependent on cotton exports. Thus, in spite of the project’s encouragement of direct demonstration to peasants and of their active participation in the technology adaptation and dissemination process, the vertical and all-encompassing character of the filière posed the sociétés cotonnières as obligatory points of passage . Indeed, they are bound to become protagonists in the project as it proceeds to extend beyond research institutes into technology transfer to farmers during Phase II. Phase II is outside of the chronological scope of my fieldwork, so the cotton companies were not really central actors in the assemblage I observed.

There was however a fundamental way in which they,blueberry grow pot and the filière at large, came into the core of the project during this first phase: through the way they shaped local production systems. Local production systems came into the project’s experimental grounds, on their turn, less in terms of how peasants actually grew cotton than in terms of a working notion that was available in the research institutes and elsewhere in the partner countries: the milieu paysan.The notion of the milieu paysan, or peasant environment, indicated how cotton production was carried out in rural areas in the C-4 countries, fundamentally made up of smallholdings in customary land. This is a vital scale of context: the ultimate destination of the technologies being transferred by the project. As remarked, local production systems came into the experimental work not as actual practices but as the control situation – the témoin – against which the performance of Brazilian technologies would be tested. How did the Brazilians get at a standard working definition of the West African milieu paysan? Some basic facts and figures about cotton production systems in the C-4 countries were included in the original project, usually in comparison with equivalent Brazilian figures: average production per hectare, average availability of capital goods and labor per household, deployment of basic inputs such as fertilizers, seeds, pesticides, or growth regulating hormones. The working notion of peasant agriculture employed at the project front line did not come however from preliminary reports or hired consultancies. It was, rather, largely taken as the local partners in the project presented it. African researchers and technicians would commonly speak in terms of a generic figure of le paysan, even if they were well aware that, on the ground, peasants’ practices showed myriad variations.

This image of the milieu paysan was far from that of an autochthonous peasantry practicing age-old traditional farming methods, in need of salvation by modern science. On the contrary: it denoted people who had already been touched by modernization – that is, by previous development initiatives –, and who had not been saved by it. For the African partners, the picture was well-known: constraints to increased productivity and farmers’ incomes had been relatively well studied, but the means to overcome them were not always forthcoming or under the control of research institutes, sociétés cotonnières, or even national governments. Some of the most intractable ones mentioned by them referred to instability of world prices, increasingly irregular rainfall patterns in the Sahel, and poor state policies and lack of budget and capacity to implement them. But like Embrapa in Brazil, its counterpart institutes in West Africa were fundamentally occupied with problems at those scales in which they could potentially act, that is, those approachable through research: better cotton varieties, more efficient deployment of fertilizers and pesticides, or enhanced soil conservation techniques. As already indicated, however, the path between research, extension and farmers was far from smooth; yet, the working notion of peasant production scaled down into the experiments largely reproduced extension’s recommendations. The average picture of peasant production in the C-4 countries with which the project worked was that of a diverse landscape, quite different from the photos of massive, rectilinear commercial farms Brazilians would sometimes show in their power points .

In a typical West African peasant farm, you are likely to find a cotton field of one to three hectares, occasionally punctuated by karité or mango trees, surrounded closely by other crops – basic staple grains such as maize, sorghum, millet or fonio, and perhaps niébé , groundnuts, or sesame. Many farmers will have a couple of heads of oxen to till the soil, and half a dozen members of his extended family to sow the crops, apply mineral or organic fertilizers, do the weeding, spray pesticides, and harvest – first the cultures vivrières , then the cotton. After harvest, the plant residue is left on the soil for the animals to feed on . Cereals and other food crops are sold in local markets or purchased by the villagers’ organizations or local private firms. This is typically done immediately, for the farmer, who is the household head, needs cash in order to hire extra hands for the cotton harvest, pay school fees, or buy medicine for a sick child or relative. He will typically sell his harvest in a period of high supply, when prices are at their lowest. The purchasing price of cotton, on the other hand, is normally fixed by the cotton companies in advance of the season, so it is assumed that the farmer will use this information to estimate what and how much to grow that year. The households’ cotton harvest is then collected by the villagers’ association and passed on to the cotton companies, which will gin, process, and commercialize the fiber. A couple of months after harvest , the peasants are paid their production, after the credit for that season’s inputs is subtracted from the prix d’achat. Even when this price is established in advance, given an upgrade, and the cost of fertilizers is subsidized by the state – as may happen when companies and governments wish to encourage farmers to grow more cotton that year, or to revert a declining trend in production,hydroponic bucket the price paid to farmers tends to remain low . This picture of cotton production in the West African milieu paysan was compounded by certain assumptions about peasant behavior. It was most often seen through economistic lenses: assumptions of rational decision-making are what makes, for instance, price incentives so central to cotton policies in these countries. In the same vein, like much of the policy and academic literature , local researchers generally depicted West African peasants as being highly risk-averse. This tendency would stem less from irrational, tradition-bound behavior than from their particularly severe vulnerability to factors outside of their control. “One major difference in relation to Brazilian farmers”, one of the C-4 agronomists pondered in an interview, “is that here [in West Africa] the peasants’ threshold for sufferance is much lower”. Government policies and peasant associations’ funds provide for quite a limited safety net; this is worsened by regional environmental conditions, which allow for only one crop a year, and even that one crop may fail if the rains do not come as expected. “Their cereal harvest has to last during the entire dry season, otherwise they’ll just go hungry”, the same agronomist continued. “If there is a drought, pest, or some other fatality and the peasant loses his crop, he loses everything.” And their predicament may not end there, since it is not even evident that an increase in productivity will bring long-lasting benefits to individual farmers – on the contrary, it may bring losses, depending on where they are situated in the local power structure.

This can happen, for instance, when the right to exploit a plot of land that is producing well is reclaimed by the chief or elders precisely because it is doing well; for many, especially women, youth, migrants and other less privileged groups, there would be no incentive for investing in improved land fertility. These assumptions were regarded as rational drives behind the peasants’ resistance to technological change: their tendency would be to stick to what they already know – that is, what they think they can control – for they cannot afford the risk implicated in trying something new. Suspicions about new technologies were not appeased by the promises of increased productivity lavished upon them by researchers, technicians, and extension agents. In fact, technical expertise in and of itself did not seem to attract, much less convince, them. African front liners were virtually unanimous in affirming that the only ones capable of making a peasant farmer get interested in a new technology and change his behavior accordingly are other farmers – something the Brazilians, for their part, were not surprised to hear. And this would happen less through persuasion and explanation than through imitation and a certain dose of productive agonism. “You know, if the neighbor is using something and it’s working well, the peasant will feel challenged to also try it”, one of the local researchers explained to me. “It’s that challenge that makes him do it”. Horizontal, farmer-to-farmer links were therefore regarded, and to some extent acted upon , as a promising dissemination channel aside from the formal extension networks run by the cotton companies.This supposed inability of the peasant to see abstractly over the long run was taken as a key constraint to technology adoption, especially in the case of no-till. Fertilization practices in particular became an area where peasant behavior was vital. In the West African milieu paysan, fertilization was done using animal manure and/or fertilizers provided on credit by the cotton companies. But both the quantity and quality of this application was highly problematic from the point of view of the researchers; it was not just quantitatively deficient but qualitatively uneven, especially if compared to the highly technified processes found in Brazil and elsewhere. Most significantly, it was common practice for peasants to deviate a portion of the fertilizers received for cotton to the fields where they grew cereal crops. It was even said that, for many farmers, access to inputs for the latter was the main reason why they grew cotton in the first place. This preference for food crops did not stem only from immediate subsistence needs, since not all of it is consumed in the household. It was also linked to the more dynamic, and less controlled,character of the markets for these crops. “For cotton they have to deliver production and wait until they can get paid for it; cereals they can sell directly in the local market and get the cash right away”, one of the Burkinabe researchers explained. “It’s just reasonable that they privilege them; wouldn’t you do the same?” Cereals and other food crops were also part of peasant farmers’ risk management strategies, in case pests, drought, or disease came to differentially affect their crops. This diversification strategy also included cattle and other kinds of livestock such as goats. As remarked in the previous chapter, these were a key piece of the milieu paysan puzzle, especially as far as no-till was concerned. Attachment to livestock has been a topic in anthropological debates on African rural life, made famous in the field of development by Ferguson’s take on the so-called bovine mystique.Brazilians would occasionally make their own analogies about farmers’ behavior towards cattle. The fact that peasants may use oxen as a sort of savings account, for instance, is something they also recognized in rural Brazil. As one of the Embrapa agronomists illustrated, “I once asked a farmer back home why was he keeping ten poor milking cows when he could exchange them for two very good ones and have the same milk production or more. He replied, ‘because if my daughter gets sick and I have to sell one of them, there goes half my herd’.” But even though the cattle question was a major constraint to the project’s chief technological component, for the Brazilians and their partners it remained as such: a constraint to be grappled with, and not a problem amenable to intervention.