Abstract
This is the introductory paper to the June 2025 Outlook on Agriculture Special Issue on agricultural deskilling. As such it provides conceptual and case material to allow a broad range of agriculturally oriented readers to engage with the other seven papers in the issue. Following an exploration of different perspectives on skill, the paper introduces in broad terms the general argument about deskilling. The focus then narrows to the treatment of skill within agricultural development, before turning to the heart of the matter – agricultural deskilling. Finally, the papers in the collection are briefly introduced.
Introduction
Does new agricultural technology erode farmers’ skills? If so, what are the implications for farming, livelihoods and agricultural development?
It has been argued that (in at least some circumstances) the introduction into workplaces of mechanisation and automation, and the simplification of work routines, results in a loss of workers’ autonomy and skill. In manufacturing contexts, this has been called ‘deskilling’.
The claim that farmers could be deskilled by new agricultural technologies is not likely to sit comfortably with agricultural research and development professionals. The introduction of new technology, accompanied by training and skills, have been at the core of agricultural development theory and practice for over a century. Does the notion of deskilling, and the body of theory and evidence around it, demand a basic re-think of agricultural research and extension for development?
This paper, and the Special Issue of Outlook on Agriculture it introduces, are about the nexus of technological change, environmental change, and skills, with a particular focus on small-scale farming in the Global South. The importance of this nexus is found at the intersection of three widely accepted propositions. First, the 277 million farms of less than 5 ha in low- and lower-middle-income countries manage 63% and 10% of their nations’ and the globe's agricultural land, respectively (Lowder et al., 2021), and they make a significant contribution to global food security (Lowder et al., 2021; Ricciardi et al., 2018). Second, small-scale producers are particularly vulnerable to the effects of climate change (Cohn et al., 2017; Morton, 2007); and climate change is challenging farmers everywhere to adapt to new and uncertain conditions of production. Third, new techniques, practices, tools and technologies will be essential to agricultural transitions and sustainable transformations across all scales and regardless of which kinds of farming systems and approaches are considered, from ‘smart’ intensification to deep-green agroecology (Cassman and Grassini, 2020; Maurel and Huyghe, 2017; Vanloqueren and Baret, 2009).
In this uncertain and dynamic context, farmers will need to be adept at identifying, adapting and using both proven and new technologies, of many kinds. Being skilful, knowledgeable, autonomous, adaptable, and nimble in decision making might be expected to carry an increasing premium.
The papers in this Special Issue ask what the literature and debates around the notion of deskilling (or skill degradation) can contribute to our understanding of the future of small-scale agriculture in the Global South. To date, deskilling in agriculture has been explored by scholars of agrarian studies, anthropology, political economy, and science and technology studies (STS); but the language of deskilling has not yet made significant inroads into the literature or the discourse of professionals and practitioners whose primary concerns are the planning and promotion of agricultural development in the Global South.
The aim of this Special Issue is to help bring the deskilling literature – its insights, debates and challenges – to a wider audience. The suggestion is not that deskilling is a necessary consequence of introducing any new technology across the breadth of small-scale farming systems in the diverse contexts of the Global South. Rather, the proposition is that research and theory about deskilling in agriculture, or what Stone refers to as ‘agricultural deskilling’ (Stone, 2004), is important and should be part of the broad corpus that informs the planning, practice, and evaluation of agricultural development.
The literature on deskilling in agriculture continues to expand. The contributions to this Special Issue show that the discussion now encompasses skills and skilling (including reskilling and upskilling), knowledge, learning, decision making, production and social reproduction, empowerment and disempowerment, ecological lenses on deskilling, and the deskilling of researchers and consumers. These are just some of the threads in the growing literature. The value to agricultural development researchers and practitioners of this expanding conception of deskilling is the light it sheds on the relationships among agricultural learning and decision making, knowledge and technique, practices and technology. It is here, in a period of climate change, technological innovation and economic transformation, that the future of small-scale farming will be forged.
We point out explicitly at the outset that this collection does not seek to provide a comprehensive overview of the policy and practical literature about technical and vocational training and skills in agricultural development (for recent reviews of this type see: Collett and Gale, 2009; The World Bank, 2007; VET Africa 4.0 Collective, 2023).
This introductory paper proceeds as follows. The next section addresses the question ‘what is skill?’ with reference to four perspectives drawing principally from economics, sociology, and political economy. The focus then turns to the treatment of skill within the theory and practice of agricultural development, where the introduction of new technology and skill (or skilling) are tightly coupled. This sets the stage for a discussion of deskilling within agriculture, drawing on four cases that together illustrate different approaches and highlight different findings. Following a brief discussion, the contributions to this Special Issue are introduced.
What is skill?
Four perspectives
What is skill; what does it mean to be skilled? In everyday usage, ‘skilled’ is synonymous with ‘good’ in the sense of ‘competent’: a good farmer or a good researcher is someone recognised as being capable and effective in their farming or research. But this focus on the successful outcomes of an activity or occupation does not tell us much about skill.
Long-standing debates within sociology, anthropology, education, economics and STS have explored the nature and meaning of skill, the measurement of skill, and trends in and factors affecting the types, levels and distribution of skill. In his paper, ‘What is skill?’, Attewell (1990) distinguished four perspectives, which he termed positivist, ethnomethodological, Neo-Weberian or social constructionist and Marxist.
The positivist perspective portrays skill as an attribute of individuals or tasks, which has an objective character independent of the observer, making it amenable to quantitative measurement. Unfortunately, the positivists’ effort to quantify has been hampered by the diversity of skills and the qualitative differences among them: for example, how can physical skills be compared with mental or social skills? This qualitative diversity complicates the calibration of a ‘common yardstick’ needed to make different skills commensurate, and thus comparable in economic value.
According to Attewell, ‘at the core of the [ethnomethodological] perspective is the idea that all human activity, even the most mundane, is quite complex […] requiring complex coordination of perception, movement, and decision, a myriad of choices, and a multitude of skills’ (Attewell, 1990: 429–430). Advanced skills involve a depth of perception to recognise patterns and confidence to choose or develop appropriate responses, where ‘skill inheres in […] recognizing something new as something old, in acquired or trained “blindness” to uncertainty and uniqueness’ (Attewell, 1990: 433).
The Neo-Weberian or social constructionist perspective ‘tries to understand the conditions under which occupations are socially demarked as skilled and the processes by which some jobs come to command higher standing than others’ (Attewell, 1990: 435). Those processes might be market-based, that is, where the supply of and/or demand for skills are influenced by social action, such as through guilds, apprenticeships, unions, secrecy and so on. The socio-economic status of skilled work is also shaped by ideological claims about the value attributable to different kinds of work, reflected for example in distinctions between ‘professions’ and ‘occupations’, and between men's and women's tasks.
Marxism is Attewell's fourth perspective, and he starts by suggesting that many Marxist scholars treat skill as a ‘common sense’ category requiring no further articulation, while some neo-Marxists might ‘shade into either positivist or social constructionist thinking’ (439). Skills were nonetheless important to the Marxist analysis of commodification (of labour) and alienation (of workers from the products of their labour).
Skill as human capital
The positivist perspective on skill is evident in the economic concept of ‘human capital’, which can be thought of as the stock of habits, knowledge, skills, social and personality attributes (including creativity) embodied in an individual's ability to perform labour so as to produce economic value (Davis et al., 2021). At a macro-level, it is widely accepted among economists that the sum of a nation's human capital is a key determinant of national economic growth and development.
The concept of human capital remains central to mainstream development thought, as illustrated by initiatives such as the World Bank's Human Capital Project and its Human Capital Index (HCI) (The World Bank, 2018). Economists predict that investments in education and training will yield economic dividends in job creation, income generation, technological innovation, economic growth and human wellbeing. The positivist perspective on skill as human capital is reflected in policy rhetoric about modernisation, economic growth and structural transformation, and in talk about increasing human capital via education, training, learning, skill development, capacity building and so on.
Human capital theory assumes that skill is a personal attribute that accumulates within people over time, implicitly measurable as the sum of years of basic, higher and vocational education and training, plus years of on-the-job training and experience. Since skills are embodied within people, they can be carried by workers from one job to another, between organisations, economic sectors and countries. Wages are assumed to be proportionate to human capital, so that higher skills command higher wages as a return on investment in education, training and experience. The human capital of organisations is the sum of the skills of the individuals working for those organisations.
The human capital approach is very much in evidence in a recent FAO and IFPRI report entitled Investing in Farmers: Agriculture Human Capital Investment Strategies (Davis et al., 2021). The authors operationally define human capital ‘as the skills and capabilities of small-scale agricultural producers to successfully manage farming enterprises’ (5). The motivation for focusing on human capital is revealed by the report's emphasis on ‘disruptions’ – from climate change and COVID-19, to droughts, digital agriculture, and ‘advanced technology’ – leading to an expectation that ‘re-skilling and upskilling programmes must be part of the response to address the needs of agricultural workers facing such disruptions’ (9) (also see The World Bank, 2019).
Skill as autonomy
Perhaps the most consequential work on skill from the Marxist perspective is Harry Braverman's book Labor and Monopoly Capital, published in 1974. Braverman, like Marx, used the craft worker as the benchmark for skilled work, as ‘the combination of knowledge of materials and processes with the practiced dexterities required to carry on a specific branch of production’ (Braverman, 1974: 443). In this conception, a craft worker has autonomy and exercises control: s/he ‘decides how to accomplish a particular piece of work, chooses the appropriate tools and procedures, and is self-directed in the work’ (441). The capitalist factory system decomposed the production process into discrete tasks and imposed a division of labour that obliged individual workers to focus on one specific operation. In effect, conception (and thus control) was separated from execution, and management became a separate professional task with a skillset of its own. The linking of skill with autonomy is critical: for some who share this perspective, ‘if workers do not decide on what tools or methods to use to accomplish a task and if they cannot schedule what to do and when, they lack not only control but skill’ (442). In the factory context, the workers’ loss of oversight and control is equated with a ‘deskilling’ of work and workers (a term that Braverman himself did not use).
Form (1987) reminds us that ‘For over two centuries social scientists believed that the mechanization of labour and the factory system speeded up the division of labor, diluted workers’ skills, and increased their unhappiness’ (29). Braverman re-ignited the modern debate about deskilling, arguing that: …not only does their [workers’] skill fall in an absolute sense (in that they lose craft and traditional abilities without gaining new abilities adequate to compensate the loss), but it falls even more in a relative sense. The more science is incorporated into the labor process, the less the worker understands of the process; the more sophisticated an intellectual product the machine becomes, the less control and comprehension of the machine the worker has. (Braverman, 1998: 294)
It should be noted that Braverman's fundamental observation, and the implications he drew from it, are not universally accepted. Indeed, they are contested with regard to whether, where, when and under what conditions deskilling was and is observed, and if so, the factors driving the deskilling process (for critiques of Braverman's arguments about deskilling see Attewell, 1987; on the deskilling debate more generally see De Pleijt and Weisdorf, 2017; Form, 1987; Myles, 1988).
Skill as status
The ethnomethodological perspective leads to the recognition of an important paradox about skill. As soon as a person can do something fluently and with apparent ease, the skill itself seems to disappear. According to Attewell, ‘skill inheres in the ability to do [an activity] without thinking about it’ (Attewell, 1990: 433). This happens to different degrees for practitioners themselves and for people observing the activity. Most able-bodied adults think nothing of being able to walk down a flight of stairs, nor are they especially impressed to see another adult do the same. But this activity requires advanced motor skills that had to be learned, and with some difficulty; once learned, they became ‘second nature’ and largely vanished from sight. They became embodied within their practitioners, subconscious for the walker themselves and unremarked by most people watching.
More esoteric and unusual skills might remain conspicuous, both for the practitioner and other people. Professional dancers and barristers are well aware of their own advanced skill. This is a function of three factors: the amount of time and effort it took to acquire and sustain the skill at a high level; comparison with others working in the same field; and the existence of institutions that signal appreciation of particular skills with material rewards and cultural recognition. Nevertheless, it is easy to see why skill becoming invisible or buried would be highly problematic if, like the positivists, you want to measure skill.
The Neo-Weberian or social constructionist perspective suggests that professional or occupational status depends more on power and political bargaining than on objectively different levels of skill. A prime example is the relative status typically accorded to managerial work vs. manual work. The higher status of management has less to do with higher levels of skill, and more to do with the carefully created and maintained belief that oversight, supervision and higher responsibility for distributing group-level resources constitute higher levels of skill. In this discourse, responsibility for overseeing and coordinating the disparate parts of a production process is the counterpart of the deskilling that Marxists perceive in the plight of workers consigned to performing discrete tasks within that larger process.
Technography, which is an ethnomethodological approach to studying and analysing practitioners of technology and their skills, offers an alternative view. It sees the unequal distribution of power between management and workforce as an effect of ‘task ordering’, whereby the performance and coordination of tasks in time and space often entails an uneven distribution of information and knowledge among those performing the tasks. This enables power to accumulate to personnel occupying key positions within task groups, such as overseers, managers and the controllers of tools and resources necessary for the task. According to this interpretation, while skill appears to be a product of talent honed by training and experience, it is really an effect of task performance, which bestows privileges and status on actors who perform certain roles (Arora and Glover, 2017; Jansen and Vellema, 2011). In the agricultural realm, technography has been used to analyse a range of cases including for example oil palm processing groups in Ghana (Adjei, 2014) and precision agriculture in The Netherlands (Velden et al., 2024).
The technographic perspective on skill in small-farm contexts emphasises the real-time decision-making and performance skills of farmers and labourers, including a focus on improvisation and coping in the face of environmental, market and other uncertainties. These performance skills are embodied, practical and tacit rather than articulate. A good outcome (i.e. a skilful performance) depends on adjusting nimbly and effectively to a dynamic context. This kind of skilful improvisation combines a sensitivity to environmental signals with an ability to choose and enact a repertoire of practised techniques and routines to suit the circumstances (Flachs and Richards, 2018; Glover, 2018; Richards, 1986, 1989). Emery Roe has compared the improvisational performance skills of pastoralists with the skills of professional managers of complex industrial and infrastructural systems. In both cases, achieving successful outcomes depends on being able to vary management processes considerably to cope with varying conditions, in order to achieve outcomes that are as stable and reliable as needed (Roe et al., 1998).
Skill and agricultural development
The transfer of information and skill has been a major pre-occupation of states and development agencies in their interventions in rural areas, with agricultural extension and other forms of training considered core to agricultural development. The skills agenda is rooted in the long-standing proposition that farming must become more productive, more efficient and latterly, more sustainable. And the prevailing orthodoxy is that change along these lines requires the use of new or improved technology, practices and management, as well as new institutional and organisational arrangements, and a re-orientation on the part of small-scale producers to ‘farming as a business’ (Rock, 2023). Historically, the critical assumption underpinning this was that most smallholder farmers lack the skill(s) to make effective use of modern technologies, practices and opportunities, and must therefore be trained (that is skilled, reskilled, enskilled or upskilled) (Box 1).
An extension perspective on skills.
‘Although farmers already have a lot of knowledge about their environment and their farming system, extension can bring them other knowledge and information which they do not have. For example, knowledge about the cause of the damage to a particular crop, the general principles of pest control, or the ways in which manure and compost are broken down to provide plant nutrients are all areas of knowledge that the agent can usefully bring to farmers.
The application of such knowledge often means that the farmer has to acquire new skills of various kinds: for example, technical skills to operate unfamiliar equipment, organisational skills to manage a group project, the skill to assess the economic aspects of technical advice given, or farm management skills for keeping records and allocating the use of farm resources and equipment.
The transfer of knowledge and skills to farmers and their families is an important extension activity and the extension agent must prepare himself thoroughly. He must find out what skills or areas of knowledge are lacking among the farmers in his area, and then arrange suitable learning experiences through which the farmers can acquire them.
Source: Oakley and Garforth (1985: 11)
Training is typically seen as an essential component of development interventions to promote the uptake and scaling of crop and livestock innovations (Shilomboleni and De Plaen, 2019; Van Loon et al., 2020). The requisite training is to be provided by agricultural extension agencies, NGOs and other organisations. The skills to be transferred are often confined to the set of protocols, techniques and schedules recommended by agronomists. Deeper understanding of underlying biophysical principles or mechanisms is not usually an element of agricultural training syllabi. In the words of Davis et al. (2021: 14), ‘agriculture human capital is often invested in as a means to an end rather than an end in itself’.
To be sure, there have been changes over the years, from positivist to more constructivist approaches. These include a recognition of indigenous and/or local knowledges and skills, a willingness to engage with local expertise through participatory research methods (DeWalt, 1994; Hoffmann et al., 2007), and a shift from singular ‘package of practices’ recommendations to more flexible ‘baskets of options’, designed to enable farmers to choose if, how, and when to engage (Ronner et al., 2021). Group-based, field-oriented learning in Farmer Field Schools ideally provides opportunities to move beyond the linear transmission and implementation of training advice (Braun et al., 2006).
The skills agenda for agriculture also appears to be broadening in at least two ways. First, alongside increased attention to youth involvement and women's empowerment in agriculture, discussions about skills now often include entrepreneurial and business skills, social and interpersonal skills, digital skills, life skills, organisational skills, leadership skills and so on. Skills and skilling have moved up and out from the narrow horizons of agricultural extension and are now seen to encompass a broader system of institutions, organisations, objectives and learners – a ‘skills ecosystem’ (The World Bank, 2007; VET Africa 4.0 Collective, 2023; cf. Brown, 2022). There has been increasing attention to integrating professional, technical and livelihood skills into formal education curricula, as suggested by the terms Agricultural Education and Training (AET) and Technical and Vocational Education and Training (TVET).
Second, there are signs of evolution in how success is defined, with Davis et al. (2021: 26) suggesting, for example, that agricultural human capital investment is successful ‘when it helps generate and develop skills, knowledge, capabilities and awareness among agricultural producers and, rather than simply transferring information, technology, or assets, it empowers farmers with full ownership of the newly acquired human capital’.
Agricultural deskilling
The literature around deskilling in agriculture (and even more so in relation to agriculture in the Global South) is a small part of the very large and diverse deskilling literature, and it is by no means the main focus of the debate. Nevertheless, the idea and language of deskilling has been used in a range of agricultural settings to study the effects on farmers’ skill of technological (and other types of) change.
Several authors, including Stone (2007) in some detail, have noted that the capitalist, industrial labour context that features so prominently in Braverman's analysis is very different from what is seen in agriculture, and particularly small-scale agriculture. Another difference between industry and agriculture is the much greater significance of environmental uncertainty to farming and farm management compared to factory work and factory management. This uncertainty shapes the context within which farm skills are learned, developed and put to work. In other words, Stone's term ‘agricultural deskilling’ not only suggests the sectoral focus (agriculture as opposed to industry) but also that the processes and mechanisms of deskilling in agriculture are distinct from those described by Braverman and others in factory settings.
We note also that, in addition to work discussing deskilling in relation to farmers, there has been some use of the notion of deskilling in relation to other agricultural professionals, such as researchers (McLellan, 2021), and this is further developed in Timothy McLellan and Ben Eyre's contribution to this Special Issue. Slightly further afield, Jaffe and Gertler (2006) argued that many North American consumers no longer had the skills required to make use of basic commodities allowing them to eat a high-quality diet on a lower budget. According to these authors, consumers had also lost the knowledge to make discerning decisions about the multiple dimensions of food quality. Sarah Cramer and colleagues have studied efforts to integrate kitchen gardens into school curricula in the USA, which they saw as ‘sites of prospective resistance to the decline of food production, preparation and consumption knowledge and skills; the phenomenon broadly known as food system deskilling’ (Cramer et al., 2019: 507); Cramer develops a similar theme in this Special Issue with her contribution about learning and teaching about food production within a prison setting. Below, we focus on four cases where deskilling has been discussed specifically in relation to farming.
Four cases
Here we refer to four cases that illustrate some of the ways that the notion of deskilling has been brought into the analysis of agricultural change. The cases comprise hybrid maize in the USA, contract farming of bananas in the Eastern Caribbean, contract broiler chicken production in the USA, and transgenic cotton in India.
Over a 10-year period from the mid-1930s, hybrids replaced open-pollinated maize varieties throughout the US corn belt. Fitzgerald (1993: 327) argued that ‘hybrid corn was an agent by which farmers were effectively deskilled’. Significantly, Fitzgerald's focus was not on the replacement of manual skill by machines, nor the simplification of complex, skilled tasks by fragmentation of the labour process. Rather, she was concerned with ‘mental skill’, specifically ‘The knowledge farmers had created over the years – regarding the differences between various lines of corn, their growing range, maturing rates, value as livestock feed or on the open market, susceptibility to bugs or disease, and in particular the visible features that good corn should have for their particular farms’ (343). Fitzgerald argued that hybrid varieties, as well as the manner in which they were marketed, made that kind of know-how obsolete for farmers, as their ‘authority and knowledge were thereby delegated to geneticists and seed dealers’ (343). This left Midwestern corn farmers unable to understand some aspects of their own operations without expert help. Notably, while Fitzgerald felt that this was clearly a case of deskilling, she pointed out that, unlike many industrial examples, farmers did not struggle against the introduction of hybrid corn technology, and that corn growers generally benefited from growing the new varieties. For Fitzgerald, the value of deskilling is as a heuristic device for understanding work, which can be useful ‘even when shorn of its political charge, and can reasonably be extended to work sites off the shop floor’ (325), such as farms.
Grossman's study of the deskilling of banana contract farmers in the Eastern Caribbean (Grossman, 1998) engages more directly with the central concerns of Labour and Monopoly Capital. Grossman wrote that the study's ‘labor question’ was to investigate ‘the extent and import of control exercised by capital and the state over the peasant labor process’ (3) as a result of contract farming. The separation of conception from implementation is one of the indicators identified by Attewell (1990) as a signifier of control moving from craft producer to capitalist master. Contract farming, in general, had already been identified as an important context for the study of agricultural deskilling through its impacts on producer autonomy and control (e.g. Mann, 1990; Watts, 1994), although the category of contract farming is very diverse and heterogeneous, and must be approached with care.
Grossman found limited evidence of deskilling. First, the technology employed in banana production in the Eastern Caribbean became more complex over time, which required increasing levels of technical skill on the part of banana growers. Second, capital and the state did not control all aspects of the production process uniformly: apart from harvesting and packing, peasants retained considerable autonomy over production. Third, farmers had been obliged to readjust how and when labour was used on the farm, which was ‘an element of conception ignored in the deskilling debate in relation to contract farming’ (125).
Grossman concluded that ‘caution is warranted in uncritically adoption the concept [of deskilling]’ (25). ‘The basic problem in applying perspectives developed to analyze changes in industry to the realm of agriculture and contract farming is that the labour processes in industry and agriculture are fundamentally different, a function of the environmental rootedness of farming (209). And it is because of that environmental rootedness that ‘farming requires a much more complex set of responses from growers than assembly line processes do from deskilled factory workers; farming is not a repetitive motion exercise’ (125).
Elizabeth Miller (2018) used Braverman's work to analyse contract farming of broiler chickens under monopoly and monopsony conditions in the USA. She undertook detailed ethnographic research on two ‘industrialized contract broiler farms’, which she described as: a model where farmers agree to raise chickens for meat for a set amount of time, at a rate of pay based on the ratio of feed to chicken weight at slaughter. Farmers invest in the built infrastructure to execute this process, but the company they contract for [the integrator] is mostly in control of the upstream and downstream supply and processing chains that depend on the production of the broiler chicken for their continued functioning. (iii)
Miller identified several factors that contributed to deskilling, including the introduction of off-farm inputs and experts, corporate concentration and changes in economies of scale. She found that the integrators actually wanted farmers to be deskilled, as reducing the variability of farmers’ labour processes helped to create more certainty in the production system. Deskilling would allow them to reduce farmers’ pay because ‘so much relative value is created outside of the farm, on grandparent flock farms, in laboratories, at the further processing plant, in the marketing department, and so on’ (166).
Glenn Stone analysed the case of transgenic, insect-resistant Bt cotton in India (Stone, 2007, 2011; Stone et al., 2014). He observed that cotton farmers in different villages of Andhra Pradesh (in a locality that is now part of Telangana) were exhibiting remarkable crowding behaviour in their seed choices. In a context where many new Bt varieties were being marketed, with relatively little information about their characteristics or relative performance, highly localised seed fads saw farmers in villages just a few kilometres distant from one another planting different cotton varieties in the same season, despite very similar growing conditions. Farmers were also rapidly switching from one seed brand and variety to another, and not, as might be expected, converging over the space of a few seasons towards a common seed choice. This seemed to contradict the expectation that farmers would make rational economic calculations based on the expected agronomic performance of the available seeds. Further, Stone observed that the farmers were being bombarded with marketing and advertising, that many of them relied on seed dealers to recommend which seeds to plant, and that many farmers were mistaken about the genetics of the cotton varieties they were ‘choosing’.
From these observations, Stone concluded that in this particularly dynamic context farmers were being deskilled because of shifts in the way they were learning about the new cotton varieties (see next section on Deskilling and learning). However, he noted that ‘crop genetic modification is not inherently deskilling’ (Stone, 2007: 67). The links between learning, decision making and technology, have also been explored for other contexts and crops (Flachs and Stone, 2019; Glover et al., 2020), and are further developed in Stone's contribution to this Special Issue.
Deskilling and learning
Two developments in relation to agricultural knowledge production and learning are relevant. First, at an organisational level, ideas around innovation systems and knowledge systems emphasising interaction, networks and links among diverse actors have been brought squarely into agricultural development (Hall et al., 2003; Roling and Engel, 1991; Sumberg, 2005). This perspective highlights the presence of multiple sources of information, knowledge and innovation (Biggs, 1990), including farmers, the private sector, universities, NGOs, research institutions and extension.
Second, at the farm level, Stone (2016) set out the foundations of what he called a general theory of agricultural learning, including three modalities of learning. Stone uses environmental learning to refer to adaptive experiment-based learning, which involves ‘observing and basing decisions on empirical payoff information’ (6). This is the kind of learning that is highlighted in the literature on farmers’ experiments, but as Stone argues, there are inherent problems with environmental learning that limit its contribution to knowledge generation. Social learning involves changes to practice based on evaluation of someone's social status, rather than on payoff signals from that person's farming practices. One's choice of an individual or group from which to learn might be affected by prestige bias, conformist bias or aversion bias, and Stone draws attention to the potential for social learning on its own to be maladaptive.
While there are potential limitations to both environmental and social learning, they are nevertheless ‘relatively free of external interests’ (Stone, 2016: 10). On the other hand there is didactic learning, which Stone uses to refer to change in farmer practices resulting from the intervention of ‘agricultural didacts’. These didacts are external parties such as public agricultural extension, commercial input suppliers, NGOs and the like who, in Stone's view, act ‘primarily out of their own interests while claiming to act in the farmer's interests’ (10). The tools of didactic learning include instruction, demonstration, exhortation, advertising, regulation, coercion, adulation, and shaming (10). Agricultural didacts have been active in many rural areas in the Global South for a very long time, starting with colonisers, missionaries and early traders in commodities wanted by the metropole.
One of Stone's key propositions is that in some situations environmental learning becomes increasingly costly or inaccurate. When this happens, as in the transgenic cotton example discussed above, farmers will increasingly rely on social and/or didactic learning, with potentially significant negative consequences. Autonomy and control on the part of farmers require learning modalities that develop both the skill required to make decisions (e.g. where when, and how to do something), and the skill required to implement them.
Discussion
Evidently, deskilling can mean different things. There is no uniform or consistent story within the literature on agricultural deskilling, and given the diversity of agriculture across the globe it would be surprising if there was. At any rate, the evidence does not support a universal net loss of agricultural skill. Instead, there are more subtle and particular stories of technological change linked to changes in technical practices and skills, and redistributions of agency and power.
The value to agricultural development of the diverse and growing literature on deskilling is the light it sheds on the critical relationship between learning, decision making and technology, for it is here, especially in a period of great uncertainly, that the future of small-scale farming will be forged. This relationship is also at the heart of Stone's model of agricultural learning, which poses a pointed challenge to agricultural development professionals, who are, by definition, didacts. If, as Stone suggests, even non-commercial agricultural didacts are primarily motivated by their own interests, is it possible to imagine trajectories of technological and other types of change, and ways of introducing them, that would not reduce farmers’ autonomy and control? And if so, what might they look like?
We note that, by exploring the deskilling literature, we have ended up discussing skilling, and that term should be understood to embrace concepts such as upskilling, reskilling, and skills transformation. Skilling is an ongoing process in any socio-technical system and is especially important when those systems are changing. Attention to skills and skilling will be vital for shaping the future of small-scale agriculture and should thus be foremost in the minds of agricultural development professionals.
Skill may be conceived as an attribute of individuals, as the positivists would have it; so that skilling is essentially an individual process of learning and skills acquisition. Anthropological and ethnographic perspectives suggest that, on the contrary, skill should be conceived as a social or cultural institution; or as an emergent expression of interactions among humans and nonhumans in a particular setting (temporal, geographical, political, economic, historical…). Skill may be understood as a feature of production processes (i.e. techniques, tasks, operations, transformations, and production chains); of communities of practice (i.e. task groups, occupations, professions); or of technological cultures (i.e. social, political and economic institutions). In fact, it is linked to all three.
The literature suggests a range of conceptual tools that can help agricultural researchers and professionals grasp the specificity and embeddedness of skills in particular contexts and situations, for example, the social–ecological system (Scoones, 1999) or niche (Descheemaeker et al., 2019); the cropscape (Bray et al., 2019); or the technology complex, a concept that links agroecology and geography with technology and political economy (Pacey and Bray, 2021). Thinking along these lines, Maya Marshak and colleagues have countered Stone's concept of agricultural deskilling with an alternative, ecological deskilling in agriculture, in order to emphasise the ‘disruption of the co-evolution of knowledge between smallholder farmers and agroecological environments’ (Marshak et al., 2021: 1189). These concepts emphasise the contingency of skills that emerge from situated agroecological and social relationships.
Francophone technographers, notably the French agronomist, historian and museologist François Sigaut, have long advocated the methodological approach of observing individual operations and chaînes operatoires (operational sequences) to gain insights into techniques and skills in farming and other contexts (Sigaut, 2002). He was particularly clear that techniques and their associated technical skills were socially produced by their practitioners.
An adjacent concept is the skill ecosystem, as extended by Brown (2022) to analyse agricultural training programmes in India. It focuses on the role of contextual and institutional factors in shaping the acquisition and utilisation of skills.
We note from the above divergence in grappling with (de)skilling that it is very valuable to distinguish carefully between technique and knowledge. There is considerable debate about how to define technique and its relationship to knowledge, but we find the words of Francesca Bray to be helpful: ‘Techniques mediate dialectically between knowledge and matter’ (Bray, 2013: 179). Bray's summary suggests that we might think of practices as the domain where skills emerge and are expressed – a domain where Stone's attention to learning and decision-making can be explored.
Perhaps it is helpful to distinguish different kinds of skill that we might call know-how, know-what, know-when, know-who, and know-why. At the beginning of that chain we might think of technique in its most specific sense, as a set of gestures and manipulations to transform matter. Know-what and know-when suggest a wider concern with the skills of knowing (or deciding) what to do (which may imply making choices between alternative techniques) and when to do it. Know-who draws our attention to the skills of navigating social relationships, organising and coordinating tasks, and disposing of resources. Finally, know-why implies the domain of theory, reasoning, and propositional knowledge.
As researchers and practitioners taking an interest in small-scale agriculture, we might focus on these different kinds of skills at the levels of the task (such as sowing), the operation (e.g. raising a nursery), the sequence of operations (e.g. through a growing season), and right up to a whole cropping system or farming system.
The special issue
The papers in this Special Issue contribute to our understanding of the theoretical, conceptual and more applied aspects of deskilling and skilling in agriculture.
'Skill Talk: The Struggle over Agricultural Decision-Making' by Glenn Davis Stone, offers a personal reflection on the background to his model of agricultural learning, with its three processes of environmental, social, and didactic learning. Here the focus is on the skill of decision making, and sharply divergent interpretations of how farmers make decisions, particularly in relation to the adoption of new technology. Specifically, Stone contrasts the view of ‘industrialists’ associated with companies supplying genetically modified cotton seeds, with those of social scientists and economists.
Marcus Taylor and Suhas Bhasme in 'Towards a Theory of Agrarian Skilling: Or, Why Farmer Knowledge Does Not Stop at the Edge of the Field', seek to advance Stone's analysis of farmer learning by highlighting the social dimensions of learning in smallholder agriculture. Specifically, they focus on the importance of collective learning through networks, and the fact that because farming is a deeply social process, analyses of farmer knowledge and skilling must reflect how farmers learn to manage a range of social relationships that are essential to their livelihoods. Based on their expanded concept of agrarian skilling, these authors argue that while some agricultural deskilling is indeed ongoing in the Indian countryside, new forms of agrarian skilling exist as a parallel and compensatory process.
In ‘Skilling and Social Reproduction’, Andrew Flachs, José Becerra, Fionna Fahey and Patricia Mathu move the discussion of skill, deskilling and skilling beyond technology and production, into the realm of social reproduction. Specifically, they ask how the social and ecological conditions of farmer learning are reproduced, and how, for example, the conditions of farm work, household dynamics, ecology and political economy shape the sorts of learning that can occur. The argument is that introducing social reproduction into the discussion can illuminate the contextual stages upon which people learn and perform agricultural skill.
Trent Brown in ‘Agricultural Skill Development in Neoliberal Contexts: Deskilling, or Skills’ Subsumption?’ draws from his research on agricultural trainings within the ‘Skill India’ programme to make a theoretical contribution to the deskilling debate. Specifically, he argues that the Marxist concepts of subsumption, alienation, and hegemony may be more meaningful in capturing processes that might otherwise be simplistically summarised as ‘deskilling’. Brown suggests these conceptual tools might bring greater nuance to broader debates on agricultural deskilling, particularly in the Global South.
‘Digitalisation and Skills in Agriculture’ by Thomas Daum explores the potential impacts of digital agriculture on the knowledge and skills of farmers, farm managers and workers. Daum offers an exploratory review of literature and suggests directions for research in this emerging area. He points out that existing literature tends to highlight the potential for digital agriculture to deskill or reskill and upskill farmers, depending on whether they take a more constructivist or positivist view on knowledge, skill, and skilling. But, in principle, digital agriculture might have diverse implications for farming and skilling, depending on the design of digital tools and how they are rolled out. Among other conclusions, Daum argues that there is a need for more quantitative and mixed-method studies to supplement existing qualitative studies.
In 'The Garden as Resistance: Horticultural Skill and Human Potential Behind Bars', Sarah E. Cramer provides a reflection on the notion of and assumptions about deskilling drawing on her experience teaching a sustainable food systems course and garden programme in a Florida prison. She brings ideas from asset-based development into the discussion of deskilling and skilling, which sheds new light on the meaning of skill, skilling and learning. Despite a context characterised by extreme institutional marginalisation and disempowerment, Cramer draws inspiration from the resilience, creativity and resourcefulness of the incarcerated students.
Timothy McLellan and Ben Eyre in ‘Rethinking rigor, knowledge hierarchies, and deskilled data collectors: An agenda for skilling research in global development’ critique what they see as widespread deskilling practices within Agricultural Research for Development (AR4D), practices that undermine the value and realisation of the actual and potential skills of field officers. They argue that embedded in these practices is an ideology of deskilled data collection. Acting together, the ideology and deskilling practices severely limit the quality and scope of AR4D research, and perpetuate colonial hierarchies of knowledge. They call for greater recognition of interconnections between research methods and hierarchies within AR4D institutions and collaborations.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
