Designing a territorialized work system as the making of a milieu

2.1 Case

The case was carried out at the request of a territorial administrative body, the Baronnies Provençales Regional Nature Park. The “Baronnies Provençales” is a rural territory of 1560 km2 located in the south of France, spread over 5 valleys and 82 communes (municipalities). The aim was to revive one of the area’s ancestral productions, that of the cultivation of linden trees. The linden tree produces a flower. The flower is picked by hand and then dried for consumption in the form of herbal teas. The harvesting and sale of linden blossom has existed in the Baronnies Provençales since the end of the 19th century, but production has collapsed despite actions to provide economic support. The Regional Nature Park therefore commissioned us to conduct a project aimed at reviving this production, the objective being to contribute to the re-conception of a work system with the linden tree.

In this investigation, we used a “design case study” approach [see also 16], where the challenge is to explore design opportunities, methods and practices following an action research paradigm. Accordingly with the design methods and processes developed within activity-oriented approaches [2], we applied a two-phase research approach which: (i) empirically analyzes the given activities within a specific field of application in order to “define the problem”, and (ii) proposes innovative methods in order to design the work system. “Defining a problem” is not just a matter of understanding the difficulties [17, 18]. The idea is to document the work system in depth, to identify the problems to solve but also to define new perspectives, for example by identifying the possibilities, opportunities or resources that exist within the situation, but which are not sufficiently brought into play. Design being a goal-oriented process, the purpose of “defining the problem” is to strengthen determination regarding the future of the design process [2, 19]. The objective of the second phase is to design the work system, to turn the determination for the future into an accomplishment, by applying methods such as simulation or participation, which are properly suited to the optimization of human well-being and overall system performance. In our case, phase one took place between February and December 2019, and phase two from June 2021 to December 2022.

2.2 Methods

The first stage (“problem definition”) made use of ethnographic approaches (one of us is a doctor in anthropology) rather than an Ergonomics Work Analysis [19]. It would go beyond the scope of this article to discuss the methodological cross-fertilization between ergonomics and anthropology (see [20] for a discussion). We will only underline that in both cases, the objective is to produce knowledge using an intrinsic approach (i.e., one that seeks to grasp reality from the same angle as the subject [21, 22]). There were two main reasons why we favored ethnographic methods: we wanted (i) to understand the territory’s inhabitants’ attachment to the linden tree, and (ii) to carry out an analysis based on long-term archives (production having begun in the 19th century). Four methods were used during this first phase:

Semi-directive interviews [23], the objective of which was to collect a range of points of view on the history and current status of linden tree cultivation. These interviews covered harvesting, marketing and work on territorial valorization (tree maintenance, the link with tourist activities, etc.). We conducted eighty 2-hour interviews. The following people were interviewed: pickers active and retired, farmers and employees, (current and former), traders (current and former), beekeepers, nurserymen/women, pharmacists, technicians from the Tourist Offices and the Municipal Technical Departments.

We will come back in more detail to the approaches and methods of the second phase of design (see Section 4). We can nevertheless give an indication here of the process that was implemented.

Design began after three public meetings, each of which lasted approximately three hours and was held at the end of the day. The main idea was to shift from the definition of the problem to a vision shared by the territorial actors. These meetings involved approximately 45 people, all of whom were concerned by the issues that had been identified: pickers, distributors and processors, various territorial actors (chambers of agriculture, road maintenance services, botanists, etc.) and local elected officials.

Following these public meetings, three “working groups” were created: “tree resources and their maintenance”, “product innovation” and “territorial valorization”. Their aim was to promote share scenarios for the design of the territorialized work system. A total of six shared scenarios were tested: restorative linden tree pruning, resource inventory mapping, new production of leaf powder, promotional events, knowledge of the varieties grown and harvesting of wild seeds with which to plant new orchards.

3.1 Making a milieu through the emergence of a work system

The linden flower trade began in the Baronnies in the late 19^th century when aromatic plant sellers began to organize the picking of the linden flower, and this picking became an integral part of small-scale farming based on the complementary nature of the diverse productions.

Baronnies linden flowers were distributed throughout France, respecting visual quality criteria that were all the stricter for it being sold in bulk (in 5 kg or 10 kg boxes). This standard required the flowers to have been picked when fully mature, with long, unbroken, blemish-free, yellow bracts. It is not easy to achieve this standard: the period of full bloom is short (no more than three days for a mature tree) and once harvested the flowers are fragile and vulnerable to rain, wind, excessive cold or heat, and insects. Pickers therefore had to develop specific picking and drying techniques.

The period of bloom being short, the pickers would use two techniques for extending the picking period. Firstly, they organized “harvesting itineraries,” with plantations being created at different altitudes to delay the flowering of trees higher up. Secondly, varietal selection made it possible to create varieties that bloomed early or late. Grafting led to dozens of local varieties which increased productivity by making harvesting easier (due to the way the flowers and leaves were positioned on the branch) and by improving the drying process (through more or less rapid water loss). Productivity was also facilitated by the development of tree-pruning techniques that helped the sun to reach the bracts, and by keeping the trees at a lower height, thus making it easier to set up ladders. To such an extent that in the 1950s the Baronnies was the only region in France producing linden flowers. And there were as many as 7 fairs in the Baronnies known as “linden flower fairs”.

Pickers refer to the money earned from the sale of linden flowers as a “manna” in as much as “picking requires no work” in the agricultural sense of the term: no inputs, no soil preparation, no crop rotation. When this money is not used “simply to live” it is associated with “memorable purchases”: buying one’s first moped or plans for children’s education.

However, the value of the linden tree is not simply financial. Harvesting was done not only by the pickers themselves, but also by grandparents, sons and daughters. Picking brought “everyone” together, entire families. Inhabitants of the region remember entire villages buzzing with the joy of family gatherings. Moreover, while “linden flower fairs” were places where pickers and distributors met, local inhabitants remember them as times when the inhabitants of the various valleys could get together, and as festive and cultural moments.

The value of the linden tree is not only financial. It is also emotional, social and even esthetic. The work connects the pickers to the trees, just as it brings together pickers and sellers, generations and villages beyond the hidden valleys of the semi-mountainous land of the Baronnies. The making of a milieu brings consistency to a set of heterogeneous phenomena: climatic variations, periods of unemployment, the beauty of the landscape, everyday life and projects, and the festivity of the fairs. Above and beyond the picking and sale of the flowers, the work carried out with linden trees weaves the fabric of life within the territory.

3.2 Crises and the unraveling of the milieu

Yet as from the 1970s, this milieu gradually unraveled, to the point that the “death of the linden tree” was announced at the end of the 1990s.

The first, and highly deconstructing phenomenon, was a change in cooperations between pickers and distributors. As from the 1960s, linden flowers were packed in small sachets known as “infusettes” (small infusers). In sachet form, the flowers were ground, often mixed with flowers of a lower quality. This was to lead to a decrease in the volumes of linden flowers purchased and to a fall in price. But it was also the visual quality criteria of the flowers, so important to the picking and drying techniques, that were to lose their meaning.

The second phenomenon was a reduced labor force, which occurred for two reasons. Firstly, rural desertification meant taking on employees. But French law does not allow a ladder to be an employee’s permanent work station (which would be the case here). Thirdly, the development within the territory of fruit tree monocultures was to create a significant workload at the same time as the linden trees were flowering.

These misalignments caused regular crises that impacted flower purchasing prices. But it was also the fairs that disappeared, preventing any cooperation between pickers and distributors, or any family gatherings under the linden trees. Furthermore, the landscape deteriorated due to inadequate maintenance of the trees, which sometimes had to be cut down to prevent accidents. Actions to relaunch linden tree cultivation by increasing prices were unable to prevent the milieu from unraveling. One after another the fairs shut down, picking was no longer a family celebration, and the trees deteriorated at the expense of the landscape. Filled with bitterness, the pickers abandoned linden tree flowers: the linden tree, “is worthless now.” This “worthlessness” referred to the devaluation of linden tree work, which no longer allowed people to hold together the many aspects of their existence. What died, was not just a work system, but also the benefits that linden tree cultivation provided to the inhabitants of the Baronnies Provençales.

3.3 A territorialized work system

Through the making of a milieu, a territorialized work system was woven, and it was this territorialized work system which unraveled as from the 1970s. The notion of a territorialized work system that we propose is based on the work of three authors: Georges Canguilhem, John Dewey and Tim Ingold. Each in his own field focused on the idea that humans must make do with and deal with an environment in order to create a milieu in which they can live.

Georges Canguilhem [28] particularly emphasized a distinction between an “environment” and a “milieu”. He showed that living things do not have a relationship of submission to an environment. The environment is chaotic, disorganized, marked by variations, differences and violent fluctuations that are not particularly conducive to life. Faced with the environment, living things must therefore create a milieu. The wild varieties of the linden tree were thus modified using grafting techniques, and the geographical environment was reconfigured as a result of the tree being planted throughout the territory. “To live” writes Canguilhem, “is to transform an environment in order to turn it into a milieu and mobilize norms of life” [18, p 147].

A territorialized work system is first and foremost just such a milieu. The development of the harvesting and trade of linden flowers, along with the creation of related techniques, as the expression of this normativity which consists in creating a milieu that is necessary for existence and life in the Baronnies.

However –and this is our second point –this normativity does not involve dominating a space or fighting the environment. Turning an environment into a milieu is a creative act that above all relies on a relationship of arrangement and flexibility between the living thing and its surroundings. Constitution of the milieu engages an activity similar to what Tim Ingold calls “making” [29], and that he describes as a work of weaving together the things of the world. He defines “making” as a process where the maker joins forces with the active and vibrant matter being worked and matches phenomena and dimensions that are a priori heterogeneous. So, for example, by planting linden trees at different altitudes we can match the very brief period of flower maturity with the workload of the pickers; the trees planted at higher altitudes flower later, which allows us to spread out the harvest period. This work of matching thus concerns not only the relationship with things but also cooperation between humans. The tree-grafting and flower-pruning and drying techniques used by the pickers were matched with the quality criteria defined by the retailers. Making a milieu thus implies an activity of matching heterogeneous human and non-human elements.

A milieu does not mean a place. It means a web, a system of human and non-human elements that are woven together and made coherent. And this woven milieu is “territorialized”: its extension and its elements can be indexed, located within a geographical space.

Third idea: like Ingold, Dewey also underlines the need to bring together and adjust elements that are a priori scattered. But Dewey adds a further idea: the connections established between humans and their environments must be understood within a means/ends schema [30]. Dewey’s central idea is that human acts are performed with ends in mind –not the stones with which we build, but the houses they will become. Dewey refers to these as “ends-in-view. ”

The articulation of ends and means to constitute a whole is the third dimension of what we call making a milieu. In the case of linden tree cultivation, some of the elements that are woven and connected relate to means. This is particularly the case for everything that is related to the work done with the linden flower: the growing, pruning, drying and distribution techniques organized into a system. In this respect, we can speak of a “work system with the linden tree”. But this work system must be understood as a means to achieve ends-in-view. Linden tree picking guarantees the monetary revenues that come from sales, which pay, for example, for children’s schooling. But it also makes possible family reunions during the picking period, and the sociability of the fairs, which are one of the infrequent cultural activities that take place in rural areas such as the Baronnies Provençales. It also allows for tree maintenance, thus contributing to the aesthetic dimensions of the landscape. These end-in-views are inseparable from the work system woven with the linden tree. In this sense, a territorialized work system is based on a relationship of combination and correspondence between the work system and the utilities that contribute toward life within a territory.

We can now define what we mean by a territorialized work system: it is an environment made up of a work system that focuses on what is necessary to life within the territory. To create a milieu with the linden tree is to create a work system: that of picking and drying the linden tree flowers, transforming them and selling them. But it also involves weaving links between this work system and the territory’s needs: economic needs, of course, but also family life, cultural life, and even aesthetic dimensions. From our point of view, the territorial dimension lies in this articulation between the work system and the non-work system, in weaving a production activity with its contribution to the needs of those who live in the territory. The linden tree’s territorialized work system was a milieu of work and life: it made it possible to both work and live within the Baronnies Provençales.

4.1 Setting up a facilitating organization

As some works emphasized [31–33], ergonomists must act on the organization of the design project, design constraints not necessarily being conducive to taking the work activities of future operators or users into account. But in companies, design processes are highly organized, as promoted by the American National Standard’s PMBOK for example [34]. There is an organization whose boundaries have already been established, where there are actors whose roles are well defined (users, designers, ...) and where existing rhythms and places facilitate meetings and discussions. Such an organization was totally lacking in our case. So, we needed to set up a facilitating organization and to frame the design process for the making of a new milieu. In our case, there were at least four dimensions to consider:

Sharing the definition of the problem in order to orientate the design. When designing in companies, ergonomists dialogue with the decision makers (the company directors) and the designers who make the design choices. But when designing a territorialized work system, there are no such institutional decision makers. The definition of the problem needs to be shared between a variety of stakeholders and the future objective for framing the design process has to be defined. It is for this reason that we decided to hold public meetings. The participants agreed on the idea that the cultivation of linden trees was central to a revival, due to the benefits it can bring back to life within the territory. As previously argued, design is a goal-oriented process, and there was the need to build a shared determination regarding the future. Collectively, the stakeholders came to an agreement on the ultimate objective: “to continue to live in the heart of the Baronnies Provençales, rather than going to live in the city”.

4.2 Supporting the systemic dimension of making a milieu

The second dimension relates to the systemic dimension of making a milieu, where the whole (the milieu) is more than the sum of its parts due to the need for coherence. Mendes and colleagues [35] have discussed this systemic dimension of a work system during its design by highlighting a process they describe as a “systemic appropriation”: a modification that is a priori very local spreads throughout the whole work system due to the interdependency between actors. As a result, the entire system is reconstructed from one person to the next, due to the cooperation between actors.

The process of making a milieu does not escape this dimension. Let us take an example from the “product innovation” group. Linden blossom is traditionally used for herbal teas. But our analyses had identified an opportunity to create new products for food purposes, in this case “linden blossom flour”. The idea was to add value to the linden tree leaves that result from the picking (given that only the flowers are currently marketed). An initial test was carried out in 3 farms: the branches were pruned and dried, and the leaves were then sorted and ground. This “flour” was then tested in the kitchen by students from a school of culinary arts. The experiment revealed that the “flour” should be considered as a condiment. But the tests raised questions concerning the fineness of the grind, the use of whole leaves and the optimal season for harvesting the leaves. A new experiment was therefore conducted the following season, involving former pickers. The idea was to use their knowledge of the different drying and sorting techniques. This second experiment made it possible to optimize the techniques and to improve quality. The powder obtained was passed on to pastry and cookie makers who experimented with new recipes. This second experiment validated certain taste combinations, highlighted the olfactory utility of the powder, and brought to light a possible use in cosmetics. This possibility was then tested in a third phase. It involved performers and soap makers. This experiment led to changes in the distillation techniques, and in return asked questions of the picking techniques.

Such a process has two characteristics, which were also present in the systemic appropriation process described by Mendes and colleagues [35]. The first is that it is an emergent process for the construction of the experience and the development of new ways of working. As highlighted by Bittencourt et al. [36], discussing different perspectives of design options also means bringing possible work changes into the discussion. But one needs “to experiment” in the sense of experiencing these work changes. The second characteristic is that the making of a milieu is an expanding and open-ended process, where the cooperation which is to take place is revealed gradually, as the milieu is woven. Such a process is “expanding” in the sense it is inclusive: an increasing number of stakeholders are enrolled, contributing toward the making, aligning people with a share direction, and reciprocally adjusting their ways of thinking and doing, both individually and collectively.

We hypothesize that such a developmental and expanding emergent process requires appropriate methods. In our case, we used a method that we call “step-by-step design” [37, 38]. Step-by-step design is a method where the design is based on “concrete situations” (“steps”) which are specifically defined so that the actors can develop experience by doing shared scenarios and looking for possible solutions. These concrete situations experienced by the stakeholders must lead to reflexive assessment dynamics in order to define and to perform the following step. The experience acquired during a step constitutes a substrate that makes it possible to assess and adjust the experimented scenario, but also to bring out new scenario and to associate new actors with whom to cooperate during a next step.

4.3 Designing for the community

Supporting the making of a milieu and facilitating cooperation between different stakeholders in terms of share scenarios and potential solutions does not mean that innovative solutions designed by institutional designers are not necessary. On the contrary, design for the community is needed.

Let us take tree resources as an example. The resumption of picking means having access to a sufficient quantity of trees to ensure a significant harvest. But those who wish to resume their picking activity do not always have enough trees on their land. Consequently, those who are not tree-owners are looking for non-financial solutions of reciprocity with tree-owners: the pruning that is required for picking constitutes a service provided to the tree-owner in exchange for the right to keep the flowers. Such a non-financial solution is also of interest to tree-owners. Additionally, the pickers try to create a “dispersed linden-tree orchard,” so as to perform tree rotations that allow them to manage their workload. But they must then cooperate with tree-owners all over the territory. The problem is that there is no inventory of trees in the territory, and that pickers and tree-owners (who are often tourists who own second homes in the region) do not know one another. To resolve this problem, we decided to conduct a census of trees available within the territory, and develop a computer application (see Fig. 1). This application offers an interactive map which identify potential owners and pickers, shows the trees available for harvesting, and provides contextual elements that meet the pickers’ requirements and constraints (state of the trees, distance from roads thus meeting linden flower trade criteria, presence of nearby drying locations, ...). This interactive map makes it possible to support meeting and cooperation between tree owners and potential pickers.

In designing such a tool, we design for the community. We support the making of a milieu by tooling cooperation between the stakeholders. Designing for the community is a process where the design (choice of tools, conceptual design) is based on, and proceeds from the problems and needs experienced by the stakeholders.

OPEN IN VIEWER

Fig. 1

Screenshot of the interactive map. It shows the description of a selected lime tree. Two entries can be selected: a portion of the area and the cultivars. The map then provides the number of trees available in the selected area, items that document their status, the date of their last pruning and the conditions of access, a photo when possible and finally the precise location of each of the trees concerned. Available from: http://sit.pnrpaca.org/baronnies-provencales-inventaire-tilleul/index.html.

As previously discussed, a territorialized work system questions the relationship between a work system on the one hand, and dimensions located outside work and which concern life within a territory on the other. In such a context, designing a territorialized work system leads one to examine questions relating both to work activity, and to issues outside work.

Two types of population have become involved in picking: former inhabitants of the region and “neo-pickers”. “Neo-pickers”’ are newcomers to the territory, often former city-dwellers. These people are clearly looking for alternative “lifestyles” to those available in towns. For former inhabitants of the region, the revival of linden tree picking is first and foremost a reconsideration of a certain model of agricultural production (the “industrial” model, essentially focused on single-crop farming), which has its origins not only in the fragilities that this production model presents in the face of increasingly frequent climatic variation, but also in the desire to develop a different relationship with living things (plants or animals) and in an aspiration for “other rhythms of life”. Both pickers therefore postponed certain changes to the work system due to the impacts they might have outside work, making a balance between work and outside work.

Hence the group that dealt with “tree resource and maintenance” included an aromatic plant specialist. He suggested developing new linden tree cultivars of a low height (approximately 1.5 meters instead of the current height of approximately 7 meters). There were many advantages to this solution. It became possible to plant orchards, which in turn optimized cultivation. And it avoided pickers having to spend long hours working at the tops of the trees, thus improving working conditions given the risk of falling. The notion of low height linden tree cultivation was however rejected. First of all, it was deemed to not respect the esthetic aspects of the landscape. Secondly, this solution required significant investment in agricultural equipment (both for the plantations and for orchard maintenance). But the amortizations would impose rhythms that impact outside work that the pickers (new and old) have considered as being unacceptable.

In this example, it is clear that the workers themselves balance working conditions with living conditions within the territory. But the living conditions within the territory must not be taken to mean “contextual constraints” or “resistances to change”. On the contrary, the needs of the inhabitants of the territory and their quality of life are ends-in-view in the sense of John Dewey. They are fully part of a project understood as a desire for the future, and they contribute to the stakeholders’ involvement within the design process. For these reasons, well-being and life concerns outside work and within the territory must be considered as something to bring about through design.