The experimental dispositif therefore commanded how actants were to be organized and made to relate to each other within well-defined limits of space and time. This was done so that vitalities within its scope were channeled according to a predetermined notion of what desirable elements were to be made to flourish, and what undesirable ones were to be allowed to wither away or be forcibly diverged to its exterior. Incidentally or not, dispositif is the same word Foucault chose to name the apparatus through which power flows in society. Like most of those that followed on his wake, in his oeuvre Foucault was concerned mostly with the government of human life. The experimental dispositif is also a manipulation of relations of forces, but of a non-human kind. Other than that, the similarities are remarkable. Like the Foucauldian dispositif, it is sustained by a particular kind of knowledge, manifested in the three expertises concerned here. This manipulation is also strategic, in the sense that it had a clear direction: what it did was to translate normative parameters that were external to the experimental domain into a technical, either quantifiable or discreet , form. These parameters came from the upper scales of context previously outlined: from global trade to the filière, to assumptions about the peasant environment, these were scaled down, at the micro level of the dispositif, into a central concern with yields as the core experimental parameter. Even conservation – as explained in Chapter 4,nft hydroponic the chief purpose of no-till – becomes important inasmuch as it helps sustain productivity in the long run, and does not significantly undermine it in the short run.
The experiments sought to strike a balance between short-term, economistic demands for high productivity, and long-term demands for soil conservation. In all three project components, brought together into the dispositif’s nested structure, experimental adaptation ultimately converged towards this end: increased yields became the ultimate sign that the system was working and thriving in its new environment. Thus, in their experiments, entomologists were chiefly concerned with mapping out and anticipating potentially significant pests – that is, insects that could cause sudden and sharp decrease in yields . In the project parcel, standing just above the layer of cotton at regular intervals there were white triangular boxes; these were insect traps, the only visible artifact of the pest control component in the parcel. The boxes, brought from Brazil, sheltered cards smeared with pheromone for attracting male insects, meant for a survey of insect populations in the area – a preliminary work to devising a biological pest control strategy appropriate to local conditions. During the first year, as the Malian entomologist explained, “there were observations, not results. For scientific results, it is necessary that a statistic dispositif be put in place”. The entomology dispositif proper was assembled in 2011 on the vitrine simultaneously with the varietal assay. In this trial, technicians observed and registered the magnitude and kind of damage inflicted to the different cotton varieties so that the behavior of the Brazilian ones would be compared to their control counterparts across two treatments: one consisting in zero chemical control, the other consisting in pesticide spraying according to recommendations for the milieu paysan. In the varietal test on the vitrine, yield, measured as a projected amount of kilograms per hectare, was the main parameter for comparing the ten cotton varieties introduced by Embrapa among themselves and to their local counterparts .
In the first season, productivity was also the main criterion according to which the head breeder from the Malian institute selected the two best performing Embrapa varieties to move further into the no-till fields the following year, where they would join the local varieties acting as controls. Performance of the different cotton varieties was also observed and compared according to other factors: level of damage by insects and disease, plant development , and the technological criteria listed above. In these evaluations as in genetic improvement research at large, breeders strived to strike a compromise between competing external demands on the cotton plant: farmers preferred bolls with heavier grain, as they got paid by kilogram of unprocessed cotton; ginners were concerned with the proportion of lint in seed cotton weight; the textile industry had its own standards of quality and color. Whiteness and brilliance of the fiber, for instance, were emerging from the trials as traits with the potential to improve local cotton varieties through cross-breeding with the Brazilian ones. But these and other parameters only became important provided that the concerned variety was also productive compared to the others. Similarly, in the no-till trial, most measurements were geared towards evaluating the impact of this new crop management system on the yields of different cotton varieties and the cereals grown in association or rotation with it, as compared to the control situation . Thus, the dispositif aimed at figuring out the best possible balance between biomass and yields: while cover crops intercalated with main crops were expected to produce as much biomass as possible in order to keep the soil covered during fallow, they also competed for nutrients, water and sunlight with the main crop, therefore potentially impacting its yields.
This trial was therefore principally concerned with estimating the best sowing date for each combination of plants: one treatment prescribed sowing the cover crop on the same day as the main crop, the other fourteen days afterwards. Other factors measured in this experiment addressed physic-chemical characteristics of the soil, and indexes of plant vitality such as leaf analysis, plant height and density, and germination. Yields and productivity, the focus of the economy of vitalities managed by the dispositif, depended most fundamentally on the right amount and quality of nutrients to feed the crop plants . In this respect, agronomists explicitly deployed in their experimental work a suggestive foreign trade-like idiom of nutrient export and import. Export is a normal part of agriculture, as much of the biomass that is nurtured inside the fields leaves them along with the harvested crop; the challenge is to make sure that these lost nutrients are replaced by equivalent import, in the form of mineral or organic fertilizers. Chronic soil degradation becomes a problem when more nutrients are being exported than imported over the long run. In West Africa, as we saw, not only was there excessive export of nutrients through leaching, runoff and the cattle that feeds on crop residues, but there was far from sufficient import of nutrients through fertilizers. As a result, peasant cotton farms developed, so to speak, a serious “trade deficit” in nutrients with their surrounding environment. The inevitable outcome is reduced productivity. As with the territory of nation-states, this idiom was predicated on the delimitation of a bounded space in-and-out of which elements travel: not just the nutrients,nft system but actants that carry them or dis/enable their action such as water, insects, weeds, and the crops themselves. Much like in customs and immigration, this movement had to be registered, quantified and controlled . As represented schematically in the picture below, this bounded space was geometrically delimited and internally organized according to the dispositif: size and form of the overall area , number and arrangement of internal sub-divisions , distance between blocks , number of lines in each sub-block , number of plants per line . Import of nutrients through inorganic fertilizers focused on primary minerals indispensible to plant development. In the entire parcel, these were typically brought in along the lines of the complexe coton distributed in the milieu paysan: the elementary tryad NPK plus sulphur and borum, followed by a second dose of nitrogen through urea. Part of these nutrients left the parcel in the form of harvest – along with cotton bolls, maize ears, or sorghum heads. The rest of the biomass was measured by cutting out from the soil and weighing a square meter sample; it was then left on the fields, to be reabsorbed and made available for the next season’s crops.
Measured and registered every day on a black board in a common area in the research institute, rainfall was an uncontrolled variable – in fact, the only significant one – that the project fields fully shared with peasant farms. Rains commanded the opening of the season in both sites, since only after a first significant stint of rain has fallen it is possible to sow. In the project parcels, however, sowing did not always occur simultaneously with peasant farms: at points, cotton and cover crop seeds did not arrive from Brazil in time due to the ever-present bureaucratic hurdles discussed in the previous chapter. This did not compromise the realization of the experiments, but recombined the dispositif in terms of a sowing date different from the one practiced in peasant land. Unplanned constraints could be turned into useful trials, since the sowing date to be recommended to farmers was itself one of the problems local researchers had been struggling with, given the growing unpredictability of rain patterns in the region. Towards the end of Phase I, this date was purposefully anticipated in order to demonstrate how no-till dispensed with waiting for a significant stint of rain in order to be able to till and then sow. In testing how the new technologies responded to local rainfall, researchers were gradually adjusting the tests so that an appropriate combination of variables – seed varieties, sowing date and depth, spacing, combinations of crops – would make up as best as possible for the irregular rain patterns found in the West Africa savannahs. Insects were another kind of external actant that would come and go freely. The entomologists’ protocol involved observing and comparing the damage inflicted on the different cotton varieties according to two treatments: spraying on the calendar, as recommended to the milieu paysan, and no spraying. Like their fellow agronomists, the work of entomologists was ultimately concerned with controlling the flow of nutrients. But more than the foreign trade idiom found in the soils component, or the kinship framing typical of breeding science, the entomologists’ task was most often conveyed according to a militaristic language of warfare: to “defend” crops from “natural enemies” through different “fight methods” . It basically involved protecting the crops from external threat: to make sure their vitalities were not significantly exported, or diverted away from the cotton plants, by insects and other undesirable actants such as viruses, bacteria, or fungi. Other such entities – the nitrogen-fixing rhizobium bacteria, or insects that could act as natural enemies in biological control – were being recruited in the opposite direction: into the experimental effort. Therefore, whether these and other actants would be allies or enemies, depended on their relation with the target plants and on their effects in the economy of vitalities configured by the experiments. Plants other than the main crops such as weeds were discouraged or removed, while cooperative species such as cover crops were nurtured in a controlled manner. Even water and the mineral nutrients themselves could turn from friend to foe, if present in excess: they could cause, for instance, runoff or toxicity. In all cases, the ultimate parameter was their effects on crop productivity. Besides mediating transactions with the surrounding environment, internally to the parcel the dispositif organized a whole domestic economy of nutrients and vitalities. Spatialization prescriptions based on statistic models oriented how plants were supposed to be distributed in the area, linearly organized by forced settlement and displacement schemes of intercropping and alternate sowing. Spatialization was not just horizontal; researchers also looked at the plot according to a vertical axis where cotton and the other crops stood between two domains, one subterranean and one above the ground. The plants mediated much of the traffic between them, and were themselves regarded as segmented in terms of subterranean and aerial parts . For no-till, as remarked, what happened under the ground was as much or more important as what happened above it: soil was the ultimate repository of nutrients responsible for bringing about higher yields. A chief concern was to make sure that fertilizers were effectively captured by the roots and utilized by the rest of the plant so that they would not be “wasted” . Soil was itself regarded as a segmented domain: mineral elements accumulate and move differently across various its strata; while some tend to stay in place where it is applied, others are more mobile and tend to sink into greater depths.