the traditional actors of innovation (companies, laboratories, etc.). In a way, they reactivate a new incarnation of the heroic figure (the lead users, the communities of practitioners resemble the “disinterested enthusiast” of the mythology of innovation analyzed by (Callon 1994)), which displaces, rather than overcomes, the emphasis placed on a central actor in the innovation process. The role and objectives of the users appear to be quite heterogeneous, just like the forms of relations that are established with the manufacturers and their representatives. Some of the innovations described are in fact similar to “simple” customizations, which are, after all, quite classic in the phase of appropriation of serial equipment or material. Open innovation is certainly part of a more egalitarian dyadic relationship between companies and users, but it is not a miracle solution. There are many consumers and their needs are evolving. What’s more, listening too much to certain users can destroy value for other users and even other stakeholders. Innovation is a collective activity that requires going beyond the organization-consumer dyad to take into account the wider network of actors who will support it, as well as the many material and technical elements that will influence the fate of a new product.
1.3. The socio-technical approach to innovation: networks and attachment
The socio-technical analysis of innovations makes it possible to enrich the previous contributions, in particular by going beyond the overly pronounced focus on certain components of the systems: the entrepreneur, the technical object, the user, etc. According to this relational approach, the success of an innovation depends above all on the progressive construction of a network of stakeholders who will support it and give it substance. It is therefore understandable that the central issue becomes recruiting allies, identifying their expectations and translating the project in such a way as to interest them (Akrich et al. 1988a, 2006). The originality of this point of view is summed up in the following statement, which is rather iconoclastic with regard to the usual sacralization of the inventor: the fate of an innovation does not depend so much on the intrinsic qualities of the idea or the object conceived as on the solidity and breadth of the chain that will support it. Innovating therefore consists of building and maintaining a chain of association that is increasingly extended, solid and stable, by attracting and recruiting new actors. Kline and Rosenberg (1986) develop a vision of innovation as an interactive process, or chain-linked model, compatible with the socio-technical approach. The consensus that has gradually taken hold in the academic sphere around this collective and systemic understanding of innovation has not prevented institutions responsible for innovation policies from maintaining approaches that are too linear and/or focused on a few key actors (Joly 2019).
In order to develop and strengthen the innovation network, one must regularly agree to transform the project into a new form acceptable to new entrants. The adoption of an innovation thus goes hand in hand with an adaptation, or even a reinvention of the “product” (which (Gaglio 2011) summarized through the neologism “adaptation”). Moreover, recruiting or losing an actor leads to a new network, which is likely to reconfigure the project. At each stage, “the innovation is transformed, redefining its properties and its public” (Akrich et al. 1988b, p. 31). This approach does not prejudge the decisive role of any one actor (who may be quite ordinary: a prototypist, a salesperson, a supplier, a client, etc.), especially since his or her influence may vary considerably from one stage to another. Moreover, many innovation trajectories develop despite the exit of the inventor’s network or of a key player from the beginning.
It is thus necessary to avoid the trap of reconstruction in the form of a success story, with its classic ingredients: passionate and determined innovators; their promising intuitions stubbornly propagated toward a demand (initially reticent, then benevolent); a concept or product that is “already there” that only needs to be refined to overcome technical difficulties or customer reticence, etc. However, even to understand a posteriori a success story, one must try to refuse a finalistic explanation: the receiving society, the convinced market, the controlled efficiency or profitability. “It is impossible to use the end of the story to explain its beginning and its course” (Latour et al. 1991, p. 462). It is therefore a question of starting again from the beginning of the story, in a pragmatic way, in order to describe and understand its extensions, its reversals and its adhesions; that is to say, “to explain its elaboration without assuming it to be acquired” (Latour and Callon 1990, p. 23). A first principle of symmetry follows from this: to consider the innovation under construction, without prejudging its success or failure (which must be explained in the same way), which Trabal (1999) has underlined the importance of in the sports sector. Hence the interest of innovation narratives in process studies (innovation in the making) (Hoholm and Araujo 2011), a trend that invites us to look at trajectories that are still unstable, or in the process of stabilization, rather than at the already stabilized products and the formalized collectives that underlie them. Hence the importance of works that are interested in this way in innovations that have not met their market or penetrated society (Latour 1992). This perspective can be compared with the contributions of Latour (1989), who very early on was sensitive to science in action, that is to say in the process of being made. It is indeed a way of capturing and then showing the experiences and actions of all stakeholders in the face of opportunities, uncertainties, disagreements and trade-offs to be made (Hoholm and Araujo 2011).
In innovation studies, materiality is generally taken into account, but in a somewhat reductive way. To put it simply, technical determinism makes it an obstacle (to be overcome or bypassed), the techno-centric approach focuses on functionality (to be domesticated), and diffusionism considers material elements as static and malleable entities. A second symmetry, embodied in the very notion of sociotechnics, allows the theory we are interested in to go further: the material or technical dimensions cannot be separated from the social dimensions. An innovation network is thus conceived as an assembly of human actors and non-human elements: materials, objects, prototypes, workshops, environments of use or diffusion, plans, regulatory texts, etc. On both sides, interests or constraints are redefined according to the context and concrete uses. Such a conception can seem destabilizing, but it allows us “to take the non-humans out of a status oscillating between the docile resource, mobilizable without effort by the social actors, and the absolute constraint, over which they would have no control” (Grossetti 2006).
Non-human elements are not static. They are constantly evolving (Ingold 2012): they change according to the transformations of the socio-technical network. First, because they are manipulated and transformed by the other actors of the innovation. Second, and more subtly, because their very properties are redefined according to the uses and other entities. For example, according to the phases of the innovation trajectory of kitesurfing (Boutroy et al. 2014), waves were initially “reliefs” that slowed down and disrupted an unmanageable glide, focused on speed; whereas later, new users with modified gear perceived them as “tremors”, or means of making them allies that they could collaborate with to reinvent the practice in an acrobatic mode. Symmetrically, human actors and their behaviors can thus be modified in return by the non-human elements with, or through which, they associate in the innovation.
Put another way, objects can have agentivity, that is, they have capacities to make humans act: they obstruct, incite, enable, associate, mediate (Latour 2006; Quéré 2015). For example, the Joëlette is a single-wheel supported-traction chair that allows severely disabled people to access hiking trails. Because of its shape and weight, it mobilizes and associates several conveyors and accompaniers who must be one with each other to move. The socio-technical choices stabilized in this innovation (in particular the absence of motorization) make the machine a hybrid between man and things that creates interdependence, and therefore attachment (in all senses of the term: material, social, sensory and affective) between disabled and able-bodied actors (Kasprzak and Perrin 2017).
A third principle of symmetry forces us to consider both associations (additions of functionalities, materials, actors in a network) and dissociations. Indeed, in some cases, “the structuring element is simply the removal of one of the elements” of the network (Goulet and Vinck 2012, p. 197). This is what has been observed, for example,
