Computing tools are nowadays at the heart of scientific work in both the life and social sciences, but they are also at the origin of digital services offered in many areas of daily life.
After having studied the sociology of databases a few years ago(No. 178-179), and the activity of coders(No. 206), the journal Réseaux returns to this question in a recent issue from a different angle: the study of the practices of scientists and designers of digital services. Both really produce data in the two meanings of the term," emphasizes Patrice Flichy, coordinator of this issue. "On the one hand, they make them exist, they structure them, they stabilize them, and on the other hand they present them, they show them, they make them known.
The informational device, Patrice Flichy reminds us, " is embedded in various social practices , those of the different professions that cooperate in a company, of the different scientific or technical communities, and of the decision-makers outside these communities. Designers develop, in collaboration with these various types of users, several views of the data... This embedding is complex to achieve. The results of computation or computer modeling are not directly aligned with the initial perspectives of the researchers who use these tools. The long work of alignment between various social worlds with different ways of doing things requires the construction of boundary objects that are flexible enough to adapt to the demands of each actor and at the same time robust enough to maintain a common identity.
Three articles in the dossier compiled by Réseaux are based on a close observation of very specific operations: a collaboration between archaeologists and computer scientists, a hackathon on mobility data, and the installation of digital tools in a Museum of Fine Arts.
The other two texts study two small scientific and technical communities, that of computational epidemiologists, and that of authors of U.S. patents on sleep time measurement and management.
Fabrizzio Li Vigni studies the emergence of a scientific community that has been particularly publicized with the Covid-19 epidemic: the computational epidemiologists. Most of them come from statistical physics and mobilize network theory. " Unlike classical scientific work, computational epidemiologists do not produce their data directly; their frontier position leads them to use data sets of diverse origin. Like the climatologists studied by Edwards, they spend considerable time cleaning and then homogenizing them. They can then, based on their model, simulate the computational evolution of the epidemic. Using the simulation platforms they have built, they can visualize the data with a dual objective: to test hypotheses and to make the results readable outside the scientific world, particularly by policy makers.Baptiste Kotras, Pauline de Pechpeyrou and Bernard Quinio have studied a case of experimental collaboration between prehistorians and two start-ups specializing in artificial intelligence and virtual simulations. Computational archaeology confronts two very different scientific communities. Archaeologists mobilize multiple and very specific types of knowledge that are always difficult to unify: they do not wish to formalize their knowledge "anchored in the materiality of data infrastructures". Conversely, artificial intelligence specialists need access to completely unified data, to definitive truths, which are necessary to train the machine learning algorithm.Vanessa Trupia, for her part, followed and analyzed a particular moment of creativity: a hackathon that could lead to the development of a smartphone mobility application. The data sets provided to the developers must be cleaned and then contextualized, knowing that the data was produced in another context that must be made explicit. "The hackathon thus makes it possible to open the black box of data". Through the process of exploration and cross-referencing, the data is reworked, equipped differently, to articulate the social world of transport and that of new digital services.Cédric Calvignac approaches Quantified Self (QS) through patents filed in the field of sleep management. Here, the data comes from the users. After identifying the axiological principles that are at the heart of this movement (desire for transparency through data, self-optimization, feedback loops and bio-hacking), Cédric Calvignac shows how the authors of patents are inspired by this imaginary (which he describes as the "dataist paradigm", following the Finnish anthropologists Minna Ruckenstein and Mika Pantzar) and strive to achieve it.Thomas Faugeras, for his part, analyzes a digitization project in a Museum of Fine Arts and the difficulties encountered in aligning the representations of the professionals and those of the service provider. Museum professionals are thinking of art historians (very high definition images), teachers preparing tours, and ordinary visitors (digital audio guides offering several types of itineraries). The city council wants to improve the image of its city. As for the service provider, it wants to develop a standard ticketing software. "These actors not only have different representations of technical devices, but also of target users. The long work of alignment goes through the elaboration of the specifications, the development of an intermediate object which takes the form of a model of the site, then of the final site". At each stage, components of the project are abandoned. "We end up with compromises, an "averaged figure of the user", not a new production, but a "replication" of a standard formula. The service provider takes into account the requirements of the city council, which finances and integrates the documentary material provided by the museum.Contents
- Patrice Flichy: Producing data, at the border of several social worlds
- Fabrizio Li Vigni: Computational epidemiology in the era of COVID-19. Disciplinary and political issues of a specialty based on the study of networks
- Baptiste Kotras, Pauline de Pechpeyrou, Bernard Quinio : Clues, algorithms and hunter-gatherers. Hybridization and friction of epistemologies in a computational archaeology device
- Dilara Vanessa Trupia: Open transport data and mobility application development. A work of equipment at the border of various worlds
- Cédric Calvignac: Sociotechnical translations of the axiological principles of the quantified self. Analysis of a corpus of US patents dedicated to the measurement and management of sleep
- Thomas Faugeras: A cabinet of designs. User figures at the heart of the design of digital mediation devices in a fine arts museum
Référence :
