Adjustments to the productivist model

9The 1960s saw the French agricultural sector engaged in a modernization of production systems, within a governance framework based on the co-management of agricultural development by the agricultural profession and the State (Muller, 2000). The productivist paradigm, uniting a belief in technical progress with a trend toward the artificialization of natural processes and environments, became the guiding idea of agricultural professionalization (Alphandéry, Bitoun and Dupont, 1989). In the wake of the economic crisis triggered by the oil-price shocks of the 1970s, however, critics of this dominant agricultural model began to make themselves heard. These voices began to propose alternative paths for development, emphasizing the need for caution with respect to destructive forms of growth (the Club of Rome). Drawing attention to the environmental cost [9] of human activities (Conference of Stockholm in 1972), they argued in favor of “a more self-sufficient and thrifty form of agriculture” for France (Poly, 1978). In the face of the economic crisis, more traditional and/or marginal production systems had proven more economically sound than standardized, indebted, productivist systems (Poly, 1978); as a result, the research and development apparatus began to pay more attention to farmers and organizations maintaining these systems. In 1982-83, a General Assembly for Agricultural Development – which sought to give a voice to farmers and to employees of agricultural development organizations – affirmed the need for a form of agricultural development that would be local, and that would be guided and supported by farmers themselves (Eizner, 1985).

10During the same period, existing alternative models began to be officially recognized (the Agricultural Orientation Act of 1980 included recognition for organic farming, for example) or at least considered (grass-based livestock production, for example). Farmers began to form new associations among themselves, independent of the standard agricultural development groups, in order to exchange ideas about their production systems and to generate the knowledge they needed to manage those systems. These initiatives were characterized by their openness to rethinking production systems as well as by their interest in redefining modes of agricultural knowledge construction and exchange by putting farmers at the center of the process.

11Thus, one can cite the example of the Center for the Study and Development of a More Autonomous Agriculture (CEDAPA), founded in Brittany in 1982 by A. Pochon to promote grass-based livestock systems (Pochon 2008); or the example of the Sustainable Agriculture Network (Réseau d’Agriculture Durable), founded in 1994 by producers belonging to two other existing networks, the Centers for Initiatives to Promote Agriculture and Rural Areas (CIVAM) and the Organic Farming Groups (GAB) (Deléage, 2004). Other groups were initiated by researchers, such as the Group for Experimentation and Research: Development and Local Action (GERDAL), established in 1983 by the social anthropologist J.-P. Darré. Established in the aftermath of the Etats généraux du développement de l’agriculture (General Assembly for Agricultural Development), GERDAL’s explicit goal was to advance alternative models of development. Still other groups were created by agricultural organizations, such as the Cattle Farmer of Tomorrow (EBD) network, founded in 1981 as an initiative of the Institut Technique de l’Elevage Bovin [10] and intended to provide better recognition of and support for a more diverse range of bovine production systems.

12At the same time, the development and promotion of agricultural products recognized by the appellation d’origine contrôlée (AOC) system, governed by what is now called the National Institute for Origin and Quality (INAO), helped preserve knowledge that have previously been ignored or swept aside, defending the coherence, the singularity, and the diversity of non-standardized production systems and products.

13On the other hand, the productivist model as implemented had created some serious problems, not only environmental but also social – the visible exclusion of those who had not “boarded the train of agricultural development.” Within the framework of the productivist system, the tendency was to respond to these problems by internalizing them, seeking to address them as technical questions or by relying on strategies of social marketing or mass communication, much in vogue at the time. A number of national communication campaigns were thus rolled out in the 1980s and 1990s that were intended to provide farmers identified by development organizations as “not reached by development” with technical information deemed locally important for the improved management of their farms (Blé Conseil, Maïs Conseil, Fourrages Mieux, Top Lait, etc.) or for the reduction of environmental impacts (Ferti-Mieux).

14In the 1990s and 2000s, critiques of the dominant agricultural model became more numerous and more diverse in origin, with a wide range of different constituencies – environmentalists, territorial organizations, hunters, consumers, back-to-the-landers, citizens’ groups, etc. – advancing their ideas (not always fully compatible) as to what agriculture should do or be (Alphandéry and Billaud, 1996). The rise in stature not only of these different critiques but also of the alternative models seeking to value agriculture’s various landscape, environmental, or social dimensions eventually led to an official recognition of the multifunctional nature of agriculture as part of the Agricultural Orientation Act of 1999 (Blanchemanche, 2000; Laurent et al., 2003). A new funding mechanism known as Contrats Territoriaux d’Exploitation (CTE) was established as a way recognizing this multi-functionality and assisting farmers in managing it (Rémy, 2000; Léger et al., 2006). [11]

15Modes of knowledge production and circulation were necessarily impacted by these changes, in at least three ways: i) non-agricultural actors, and especially environmental actors, became increasingly involved in agricultural governance; ii) “conventional” agricultural organizations found themselves compelled to recognize agriculture’s environmental dimensions, and began to hire specialists to assist them in this (Brives, 1998; Petit et al., 2011; Compagnone et al., 2013); iii) new agricultural organizations, taking up these new questions, began to appear. This three-way impact in turn led to an exchange of personnel and expertise between and among agricultural organizations and environmental organizations (Dupré, 2004).

A diversification of ecological models and the embrace of agroecology

16The need to reorient agricultural practices toward a better respect for the environment was subsequently taken up at the European institutional level, with the CAP reforms of 2003 including the creation of an “eco-conditionality” requirement for farmers receiving subsidies (Deverre and de Sainte Marie, 2008). A greater range of production modes was now recognized, counterbalancing the “uni-dimensionality of the productivist model” (Deléage, 2013b: 36) that had held sway during the preceding period. At the same time, new terms to describe these different forms of agriculture continued to appear: “high environmental value agriculture,” “ecologically intensive agriculture,” “conservation agriculture,” and, more recently, “agroecology,” all claim in various ways to achieve a reconciliation between the intensification of production and respect for the environment (Levain et al., 2015). Different terms are supported and promoted by different actors (public officials, agricultural professional organizations, farmers’ groups, etc.), and correspond to forms of agriculture that may not only differ significantly but also be in competition for public recognition and legitimacy (Ollivier and Bellon, 2013). This plethora of terms and types testifies to the ongoing struggle to define what will be recognized going forward as “good agriculture,” – since if we know now what agriculture should not do, what it should do remains controversial. Thus the major grain cooperatives, such as Terrena and Vivescia, have differentiated themselves from the national government’s agroecology project, launched in 2012, by developing their own models of ecologized agriculture, known as Ecologically Intensive Agriculture, following the work of Michel Griffon (2013), and Sustainable Agriculture, with the establishment of the Institute for Sustainable Agriculture in 2008.

17More generally speaking, businesses selling agricultural inputs (e.g., cooperatives and brokers/dealers) must now consider how to replace a portion of their input sales (and in particular the sale of crop protection products) with the sale of advisory services to guide a more targeted use of such products. The Chambers of Agriculture, already in decline due to reduced funding, and under pressure to improve their efficacy in guiding farmers toward the implementation of environmentally sound practices (Compagnone et al., 2013), are increasingly working on a contract basis with other agencies and organizations (water boards, local governments, associations, etc.) or with the State in the context of the Ecophyto agenda (Guichard et al., 2017; Compagnone and Simon, 2018).

18This situation, which is characteristic of the recent period, is paradoxical with respect to modes of integrating farmers’ knowledge: a tension now exists between neo-diffusionist and participatory approaches to development. The notion of transferring knowledge from the agricultural research and development apparatus to farmers is returning in force, after having weakened in the previous period, both as a result of the criticisms made of development and because of the arguments advanced in favor of a better recognition of farmers’ individual situations, and a better attention to the co-construction of knowledge within the advisory services relationship (Cerf and Maxime, 2006). Today, social requirements and environmental regulations are again becoming stronger, farmers find themselves compelled [12] to modify their practices so as to reduce their negative environmental impacts, and the Chambers of Agriculture are charged with assisting them to make this shift (Petit, Compagnone and Joly, 2015; Guichard et al., 2017). The problem of how to change agricultural practices is again being expressed by the heads of development organizations, researchers, businesses, and local governments – as it was in the 1970s, at the moment of agricultural modernization – in terms of farmers’ “resistance” to change, or the “road blocks that need to be removed” to get farmers to adopt new techniques (Compagnone, 2016). A competition also exists among different organizations as to which form of ecologized agriculture development efforts should favor. While large companies like the grain cooperatives seem to be leaning toward forms of neo-productivism (Wilson and Burton, 2015) or an “ecologizing” [13] productivism (as opposed to the productivism of agricultural modernization, which could be described as “artificializing”), other groups are promoting innovations that break more decisively with the productivist model.

19The situation is paradoxical in two respects. First, because the imperative to ecologize agriculture leads many professional and political leaders to assume that this ecologization can only take place via a massive diffusion of appropriate knowledge to the greatest number – but the knowledge needed with regard to “ecological” systems is not fully available, which means one must rely on the knowledge farmers already have or are in the process of developing. Second, because even where this knowledge is available, it cannot simply be applied in a standardized fashion, given that the principles of agroecology emphasize the specificities of local conditions and the agronomic intelligence of farmers to appreciate and make decisions as to the implementation of new practices. In this situation, the challenge for research and development is to foster the development of agronomic knowledge by farmers themselves and to benefit from farmers’ experiments and local findings by testing those results, expanding their validity, and sharing them with a larger audience. As a result of these tensions, we see the reappearance of language and subsidy participation requirements promoting the idea that farmers should form groups and work together, even though many farmers have evidently not waited for encouragement from public officials to organize themselves into groups for the purposes of producing relevant knowledge.

20This historical review suggests the way in which successive displacements over the past fifty years have been driven by the events, problems, and criticisms faced by the dominant mode of agricultural production. Thus, the institutionalization of agroecology may be considered today as a major transformational operator that is forcing agriculture to move out of this dominant regime. What impact is this likely to have on the ways in which agricultural knowledge is produced and exchanged?

Knowledge production in agroecology: local and specific

22Recent research on agroecology, in company with earlier work on sustainable agriculture, emphasizes above all the singular and local nature of the relevant knowledge. Some authors, for example, speak of the need to “deconstruct” official agronomic science in order to “reconstruct” an “alternative” science based on local knowledge (Kloppenburg, 1991).

23Both in the English-language and in the French-language research literature, many writers focus on transformations of what is commonly referred to as the agricultural knowledge and innovation system. [14] Researchers have demonstrated the limitations of a vertical, top-down, linear vision of how knowledge is produced and exchanged, suggesting instead, as a way of incorporating the singular and local dimensions of the situations encountered by farmers, that knowledge is constructed through and in action itself, in the interactions among farmers or between farmers and advisors or researchers (Röling, 1992 ; Darré, 1994 ; Ingram, 2008). [15] Such critiques were subsequently integrated within the development system in a fairly conventional fashion, with more emphasis placed on so-called participatory approaches.

24Organic agriculture, in particular, has constituted a “key social space” for the expression of this critique of vertical, top-down knowledge. Organic farming developed in a relatively autonomous manner relative to the dominant agricultural knowledge and innovation system, relying instead on its own networks. A comparative analysis of processes of knowledge construction among conventional farmers vs. organic farmers in the United Kingdom, for example, found that organic farmers had to “re-localize” their understanding of production processes in order to figure out how to produce more ecologically (Morgan and Murdoch, 2000). A similar situation exists with so-called minor products (rare livestock breeds, specialty crops, etc.), for which knowledge is lacking and cannot simply be transmitted to farmers by an advisor, but instead must be co-produced through interaction between farmers and an advisor, or between farmers and researchers (Girard and Navarrete, 2005). Situations like these can be said to reveal a shift from the “monoculture of scientific understanding” to an ecology of knowledge (Santos, 2011) in which diverse forms of information exist (Tisenkopfs et al., 2015).

25This perspective becomes central in the case of agroecology as defined by its most well-known theorists, including M. Altieri (2004) and S. Gliessman (1997). In their writings – which specifically reject modes of knowledge production associated with the “green revolution” – farmers’ knowledge is considered to be the most legitimate form of knowledge. “Peasant knowledge,” “eco-knowledge,” or “ethno-knowledge” are guiding terms in this work, used to define knowledge that peasants construct in a manner that “co-evolves” with the species with which they interact (Toledo, 1993). It should be noted however that such works relate primarily to contexts of subsistence farming or self-sufficient farming (mainly in Latin America), where agricultural systems are based on a highly developed, multiple use of natural resources. Moreover, the role of formal institutional training (basic or advanced) in knowledge construction for these farmers is often much less than it is in the French context.

26Despite these differences in context and terminology, the idea of grounded ecological knowledge is also present in the work of certain ethnographers. In France, a focus has been on “those who work with living things” (“travailleurs du vivant”), for whom natural processes are at the center of their activities, with human intervention seeking primarily to assist or to favor. This is shown by G. Delbos (1983) in her study of knowledge among salt producers in Brittany, [16] by M. Salmona (1994) with regard to sheep producers and vegetable growers, and by A. Moneyron (2003) in her analysis of the practices of transhumant shepherds. Among the latter, eco-knowledge is understood as practical knowledge that includes hazards, which appear as the linchpin of these farmers’ thinking from season to season rather than as an obstacle within their understanding of the environment. In contrast to technical knowledge as dispensed by agricultural advisors and teachers, eco-knowledge is not uniform or standardized, but is (re)constructed by each person through his or her interactions with the environment. [17]

27While this literature emphasizes above all the local character of ecological knowledge – empirical knowledge, born of rooted experience, restricted to a specific location – it is nevertheless also social knowledge, in the sense that the “local” is a construction that is above all relational, emerging from spatialized social relationships. As a result, local knowledge is continually being reshaped in light of social practices such as conversation, work, or social relationships; it is also charged with feeling (Raffles 2002). In France, this social dimension of local knowledge has been closely examined by J.-P. Darré in the context of GERDAL. It was in fact this question of knowledge production among farmers that led him in the 1970s to develop a method intended to allow farmers to work together as a group to identify their concerns, develop them into problems to be addressed, and then find or generate suitable knowledge to do so (Darré, 1978). The space of production, validation, and application of the knowledge in question is thus limited to the scale of the local group. Its extension to a larger number of individuals is only envisaged via a reproduction of the process in other areas, precisely because the value of the knowledge is linked to the identification of specific problems in specific material and social situations.

28Whether we refer to this as “ecological knowledge,” “peasant knowledge,” or “local know-how,” these forms of knowledge that link knowing and doing are thus above all singular, unique, enmeshed in the social relationships that are present at their production; and they are characterized by an adaptability that results from a constant interaction with the environment (Dupré, 1991). Some scholars now apply these notions to agroecology, and thus speak of “situated” or “contextualized” knowledge (Girard, 2014). Situated knowledge is “actionable” knowledge (Argyris, 1995); that is, knowledge that can be put into practice in specific situations. This perspective is reflected in the “practice turn,” characterized more broadly by an increase in the attention paid to practices in the work of knowledge construction (Gherardi, 2017).

The development and sharing of local knowledge

29For farmers and for research and development organizations alike, the singular nature of “situated” or “contextualized” knowledge raises a key question: how can such knowledge circulate beyond the context in which it has been produced? At issue is the need to: i) recognize or gain recognition for the work of knowledge production by those involved, along with the agricultural models they represent; ii) allow other individuals confronted with similar situations to benefit from this knowledge; iii) preserve the knowledge as a form of technical or cultural capital; iv) test it within processes of increasing generality so as to make it more widely useful. This question is fundamental in the case of agroecology, [18] especially given that, as we have noted, the knowledge produced by agronomic research on agroecological systems is incomplete and subject to verification, and requires moving beyond the classic diffusionist model. Accordingly, one sees an increasing emphasis within agronomic research on the use of on-farm research trials or participatory research as a means of producing knowledge in context and benefiting from farmers’ existing knowledge. In France, the Directions Régionales de l’Agriculture de l’Alimentation et de la Forêt (DRAAF, the organization that brings together regional departments of food, agriculture, and forestry) and the Chambers of Agriculture have issued numerous documents describing existing system types recognized as agroecological, thus helping to valorize knowledge production and improvement efforts already underway in these areas. These initiatives, it may be noted, were supported by the emphasis on “capitalization” proposed within the French state’s agroecology agenda of 2012.

30Two typical modes of knowledge exchange may be distinguished: knowledge exchange among farmers, and knowledge exchange that is mediated or organized by other actors. In the first category, many experiments in peer-to-peer knowledge exchange exist, both in the Global South, with initiatives like “campesino a campesino” (Rosset et al., 2011), and in the Global North, for example with projects to pair beginning farmers with established farmer-mentors (Chrétien and Daneau, 2013; Rengard, 2016). Knowledge thus remains profoundly attached or “stuck” to individuals and their experience, even as it circulates, in the sense that it will not be accessible or understandable by others except in the light of the experience it emerges from. Contextual elements thus need to be now supplied as knowledge is transmitted. This is what we observe when, as an ordinary occurrence, farmers adopt certain techniques or materials based on the personal experience and testimony of other farmers within their professional dialogue networks (Compagnone, 2014). Beyond this network of interrelationships, farmers who adopt new methods may be within “hearing range” of other potential users who do not belong to their immediate circle, but can listen to their testimony via a workshop or other forum, whether in person or online. We see this type of dynamic at work particularly in technical fields that are marginal to mainstream agricultural research, such as organic farming or conservation agriculture (Goulet and Vinck, 2012); in these fields, farmers can claim a role as leaders among groups of peers. In these contexts, knowledge drawn from experience becomes “common” knowledge to be shared rather than a private resource to be marketed and sold.

31In the second category – knowledge mediated by other actors – we can observe numerous efforts that seek to record and generalize knowledge produced in a specific context. Through a work of “scientization” (Agrawal, 2002), the goal is to “unstick” knowledge from the experience that produced it and/or the individuals who sustain it: to translate it into a form in which, at a minimum, it becomes accessible to those unfamiliar with the original context because they are too distant in time or space; or, in the best-case scenario, it takes on a more general character so that as to become utilizable in a larger context. The difficulty of such a process is that practical knowledge is “practical” precisely because of the experience that is connected to it. It is the “the small, detailed adjustments, acquired through experience and impossible to articulate in the form of principles” that underlie the success of a practice based on this type of knowledge (Agrawal, 2002: 330). The effort to mount in generality assumes a work of normalization and generalization that tends to efface such details.

Encapsulated knowledge and embodied knowledge

32If a tension appears between situated knowledge and formalized knowledge as soon as we become interested in the question of the capitalization and circulation of agroecological information, this tension is underlain by a larger division that exists between two ways of conceiving of agroecological knowledge. While thus far we have emphasized the importance of the singular and the local in this type of knowledge, an alternative to the incorporation of knowledge within individuals [19] (Callon, 2006) is its encapsulation (Langlois, 2002) within technical artifacts. We can define these processes of encapsulation and incorporation in the following manner.

33Encapsulation takes place via an effort to gather a maximum amount of data, obtained as much as possible in an automated fashion (e.g., using sensors or receivers), and then to organize and analyze those data (e.g., using algorithms) to rapidly generate a corresponding decision or action. This type of effort is characteristic of so-called precision agriculture, “digital” agriculture, or “intelligent” agriculture (associated in turn with certain versions of “agroecology”). The goal is to develop “smart” tools based on a continually increasing capacity for data capture and analysis. The “encapsulation” resides in the fact that the knowledge becomes housed within one or several technical objects, rendering it relatively immune to direct access by an operator, even as it promises to “augment” his or her knowledge capacity. A number of agricultural companies, notably the large grain cooperatives, are engaged in such efforts to “ecologize” agriculture based on the encapsulation of knowledge, developing their capacities for data collection, storage, and analysis and providing farmers with digital tools to monitor crops and assist in decision making.

34Incorporation is the reverse of encapsulation. It refers to an increase in competency on the part of individual farmers or groups of farmers. This process of incorporation enables farmers to improve their capacities for the observation and interpretation of specific conditions within their natural environment, and to define relevant actions to implement in that environment based on their objectives in terms of sustainability, production outputs, or the improvement of their systems. Incorporation enables the operator, within a regime of familiarity (Thévenot, 2006), to develop a unique intelligence and sense of a situation.

35Of course, real situations rarely correspond to one or the other of these two extremes, but instead take the form of intermediate configurations. Thus the development of precision agriculture or specific decision-making tools may be regarded skeptically as turning the farmer into little more than a supplement to the technical process or, on the contrary, celebrated as giving the farmer the powers of an “augmented human” (Claverie, 2010; Magnin, 2017), one whose capacities for monitoring changes in the environment and then taking an appropriate action are decoupled. (Such debates are not unique to agriculture at the moment.) The “steering” of natural processes, so central to the idea of the ecologization of practices (Larrère, 2002), may likewise be achieved either through the use of technologized devices (e.g., decision-making tools) or through a reliance on the experiential know-how of farmers.

36This opposition between encapsulated and incorporated knowledge, between reliance on technological tools and reliance on sensory perception and experiential knowledge, raises many interesting questions with regard to modes of governance and the use of technical systems, and with regard to the power dynamics involved in the mastery of such systems. Critical thinking on technique (following authors such as J. Ellul [1977] and I. Illich [1973]) has shown the risks of heteronomy resulting from the development of massive, centralized systems that are inflexible at the local level, versus the autonomy enabled by systems that are lighter, decentralized, and flexible at the local level. It is precisely this critique that was advanced by J.-P. Darré in his 1978 work, Freedom and Efficacy of Working Groups, (Liberté et efficacité des groupes de travail), in which he proposed that farmers could develop knowledge that would be useful to them while at the same time rendering them less dependent on outside agricultural advisors.

37Today, some groups of farmers adopt positions to oppose technologized forms of agriculture that promote processes of encapsulation. For example, a recent claim from the network InPACT, [20] a coalition of various alternative agriculture associations, argues that the adoption of expensive technologies (robots, computers, biotechnology, etc.) not only renders farmers captive and dependent but also has the effect of generating data that are subject to external appropriation – and that can go so far as to enable predictions of farmers’ behavior – rather than encouraging independent initiative. The netwok defends the notion of “peasant agriculture” – a term that has become a rallying cry for a diverse group of farmers and networks reacting against the government program launched in 2012 and its “technicist” vision of agroecology (Lamine, 2015). The form of agroecology InPACT favors goes hand-in-hand with a claim for greater autonomy for peasant farmers with respect to technical systems; a claim, in other words, for “technological sovereignty for peasants” (InPACT, 2016). This type of network, which can also be found at the international level (see Fressoli and Arond, 2015), does not seek to completely do away with agronomic research, however; rather, it asks researchers to investigate processes that will allow farmers to gain in competencies and in “capability” (Sen, 1992), and to expand the space and time available for the co-production of knowledge adapted to farmers’ situations.

38We see here the highly political dimension both of how agroecology should be defined and of the manner in which knowledge production and exchange should take place within an agroecological framework. The limitations of productivist, artificializing production systems and the need to find more sustainable methods have expanded the horizon of possibility. These limitations, and this expansion, have allowed models developed on the margins of the dominant system to come to the center as possible alternatives. But they have simultaneously compelled the dominant model to demonstrate that it can transform itself so as to disarm the criticisms that have been lodged against it.

Work performed by research and advisory organizations

41The first article in this issue, by Aurélie Cardona, Amélie Lefèvre and Sylvaine Simon, examines the work of two INRA experimental research stations. A distinctive feature of such stations is that they are territorially based and thus enmeshed in social exchanges with area farmers. They are not “indoor” laboratories producing results that are indifferent to the local setting. The authors of this piece focus on the manner in which, even within the process of scientific knowledge production, these research stations effect a slippage with respect to both how the research is done and the place allowed to future users of that research (that is, the farmers) within that process. This slippage takes place both in the social organization of the research and in the stated objectives of knowledge production. The goal becomes not to evaluate the effectiveness of a particular technique, but rather to focus on production systems as a whole and to seek to assess the relative value of different systems. Ecological processes must thus be taken into account as well as the social dynamics among the actors involved in the management of these systems.

42The inclusion of agricultural actors (i.e., farmers) within the research process testifies to researchers’ recognition of the important role farmers can play in supplying critical knowledge that can assist in advancing the research project and in producing relevant results for those actors. This reveals a specific form of the process of scientization: the exchange of information between researchers and other agricultural actors becomes a key element of the research method itself. Experiments and data collection are conducted on-farm, and farmers participate in the experimental design and evaluation. Nevertheless, a key feature of this type of scientization is that data analysis and the publication of results remain within the domain of the researchers.

43If this situation suggests a transformation of relations among different categories of actors to arrive at an apparently mutual benefit, this is not so much the case in the next article, by Soazig Di Biano, on the work of technician-salespeople employed by a major agricultural cooperative. The ecologization of agriculture in this case appears to be less positive. It takes place within the context of a commercial relationship; and it is supposed to involve all coop members. Whereas in the case of the INRA research stations, one can observe a broadening of the investigatory field and the range of possible knowledge to be developed; here, one sees instead a restriction of the range of investigation. Di Biano argues that this restriction takes place by means of the formalization of knowledge within guides and other tools produced to assist farmers in decision-making. It is also manifested in the promulgation of a uniform message for the coop agents, in the demand for traceability with regard to the advice provided to farmers, and in the recording of farmers’ practices by means of various devices.

44Nevertheless, as in the previous example, a form of ecologization is evident in the expansion of the range of topics considered within the advisory context: whereas in the past the coop’s advice was primarily focused on the plant, attention is now paid more broadly to the field and to the farm. Similarly, a greater number of biotic and abiotic factors are monitored on farms, and Decision-Making Tools (DMT) allow for the registering and analysis of a greater range of criteria. An overall understanding of the farm is developed within the advisory context, using data and indicators elaborated in the “back office” of the coop’s advisory services unit (Labarthe and Laurent, 2013), and with the goal of better responding to farmers’ concerns. This example thus illustrates the trend toward the development of tools that can encapsulate knowledge. Thus, as the article shows, we can conclude that in this situation the ecologization of agricultural systems is based on a neo-diffusionist idea in which the technician-salespeople acquire a new professional identity and a new legitimacy for their activities.

45The article by Nathalie Girard and Danièle Madga, on the formation and training of farmers in the production and/or use of knowledge, acts as a pivot for the collection. In the first place, it relates both to the issues present in the previous two articles and to a question central to the subsequent group of articles, viz. the functioning of groups of farmers with respect to knowledge production. In the second place, it presents a situation in which knowledge distribution takes a form midway between requiring a written account of experiential knowledge for a group of farmers and an ambition to rise to a higher level of generality with that knowledge for the benefit of a larger audience. The example presented relates to the management of grazing for animals on pasture. A group of researchers working for an organization providing advice and training on pasture management developed a method to help farmers analyze their grazing conditions and improve their management decisions within complex and changing natural environments. The researchers’ goal was not just to assist the farmers in expanding the number of indicators they could consider within a standard decision-making framework, but to assist them in rethinking the overall situation.

46The social device was hybrid in several respects: i) it was promoted by both researchers and development agents; ii) it was addressed to graziers involved in weakly artificialized modes of production, in various regions of France; iii) it was intended to support farmers in improving their ability to assess their specific situations, but at the same time sought to develop general knowledge by capitalizing on that experience. The latter strategy relied on experiences the farmers shared verbally but which were then assessed in scientific terms, with the advisors selecting what was to be retained from these experiences, the objective being to “get them in writing.” Local knowledge production on the one hand, the scientization of knowledge on the other, are connected via a triple movement: i) to allow and elicit narration as a way of fostering the exchange of personal experiences among the farmers and formalizing their embodied knowledge; ii) to disconnect these accounts from their place of production; and iii) to evaluate them in scientific terms as a way of validating transferable “local knowledge.” The method thus seeks to hold a process of knowledge production in tension at three levels: knowledge that is specific to each grazier; knowledge that is shared among the group of graziers participating in the project; knowledge that benefits all graziers at least to some extent.

Knowledge production by groups

47The second group of three articles focuses on the production of unique knowledge by groups of farmers involved in conservation agriculture, the use of open-pollinated corn varieties, and biodynamic viticulture, respectively. The work performed with regard to knowledge production by these groups is concomitant with a position taken with respect to power. Whether it is a question of gaining a measure of independence from the companies and organizations that structure and enforce conventional agriculture practices, or whether, in addition to this independence, farmers to engage in an institutional effort to gain recognition for these new agricultural models, knowledge production in these examples takes on a political dimension. Knowledge production establishes a discrete position from which another position – that of the dominant model – may be strategically challenged (Certeau (de), 1990), whether in the use of hybrid varieties or in the use of synthetic inputs. In this sense too, the group becomes a site for the redefinition of professional identities.

48To varying degrees, all three of these articles emphasize an understanding of living beings in which the temporal dimension is essential. Learning itself can only take place by means of a kind of daily wear and tear by which humans and non-human beings learn to live with one another; and the sensory perception of situations plays a central role. Knowledge here is highly embodied, singular, and anchored within specific situations.

49Jessica Thomas’s article on conservation agriculture describes how a group of farmers made use of a financial tool created by the Ministry for Agriculture as part of an initiative to promote agroecological innovation. For the farmers, this involved a shared effort to improve their practices by conducting experiments, but also by producing references to demonstrate the effectiveness of no-till seeding to public officials, and thus to bring about institutional change (Compagnone and Pribetich, 2017). The collective expertise they have developed has also enabled them to position themselves within the power relationships structuring the landscape of agricultural development. In this example, the process of the scientization of local knowledge is integrated into the approach, since the objective is to produce knowledge that is generalized enough to contest the dominant model. The idea of “producing references” illustrates how a model can establish standards as a way of competing with the dominant standardized model.

50In the next article, Floriane Derbez considers the learning processes of farmers engaged in the production, selection, and planting of so-called “open-pollinated” corn varieties in the place of hybrid corn. She traces how the farmers gradually came to “deconstruct” the use of hybrid corn and the larger system for hybrid seeds. Underscoring the fact that the technical object embodying the distribution of roles creates a specific type of relationship between the producers and users of knowledge, Derbez shows how the substitution of open-pollinated varieties for hybrid varieties entails a transformation of these roles and relationships. The specific quality of the technical industrial object being to render this imposition of roles invisible, its replacement with another object has the effect of revealing the true nature of the system. The activity of deconstruction – which Derbez, following M. Akrich, refers to as a “de-scription” – is thus not only made possible by but is also necessary for mastery of the new object.

51Thus we see the learning processes that take place and the ways in which farmers gain independence through the use of open-pollinated corn. We can observe that the process is also guided by a kind of ideology in the sense of P. Ricœur (1986); that is, by a type of thought that allows one to define what things are, to engage in a critique, and to guide action in a desired direction. The re-conquest of autonomy is undertaken specifically with regard to dominant socio-technical systems and constitutes a goal in itself. The process of knowledge production and circulation here too thus embodies a political undertaking. On the one hand, as Derbez points out, no one can hold a monopoly of expertise on open-pollinated corn; such expertise is necessarily distributed among the actors. On the other hand, the political dimension is likewise expressed in the desire to share the knowledge that has been produced more widely. This enlargement leads both to a work of scientization, enabling a passage from the writing up of unique experiences to the production of data; and a work of regulation, for a larger group of farmers. Within the process of this enlargement the group will thus make use of two registers of knowledge: knowledge formatted by science, and knowledge emerging through a sharing of direct experiences among peers.

52The article by Jean Foyer on biodynamic viticulture makes it possible for readers unfamiliar with agriculture to better understand the philosophy underlying this type of production, the forms of farmers’ engagement in this type of production, and the insights offered by an anthropological view of this type of production. Considered with respect to the types of knowledge explored previously, the author shows the vast range of forms of knowledge that can go beyond standard oppositions between situated knowledge vs. generic knowledge, or peasant knowledge vs. expert knowledge. Next to scientific knowledge, we can thus recognize what may be termed peri-scientific knowledge, experiential knowledge, sensory knowledge, even super-sensory knowledge, all of which are then gathered together in a syncretic fashion. In biodynamic viticulture, the farmer is described as being engaged as a producer in all of his or her human dimensions (including sensory, psychological, and spiritual). We see here an extreme situationism in which the total connection between the individual and his or her natural environment is engaged within the activity of agricultural production.

53But while farmers’ direct experience with things is presented as essential, we can also see that the social context plays a central role in determining access to biodynamic thought, in providing support during the conversion process, and in collective learning and the production of knowledge. Here we find ourselves confronted with true communities of practice (Wenger, 1998). The question of the scientization of knowledge arises here too, but instead of being undertaken with respect to standard science, this takes place within a contestation of standard science and a valorizing, in its place, of what we might call (following Kuhn [2008]) extraordinary science.