Pr Dominique Boullier, French digital sociologist, will be in Tirana at the end of April to give two lectures on social media (29 and 30/04). He’s currently setting up a multidisciplinary team to work on an experimental Proof of Concept to explore the dynamics of propagation in digital environments. The project will bring together data scientists, developers and designers to design a tool for monitoring the transformation of memes or viral topics, combining computational analysis and critical reflection on the effects of virality.
We are organising a meeting on the 30 April in Tirana at 6PM at the D:4 Architecture and Coworking to discuss this project with interested parties. Anyone interested should write to me at hello+event@privacydream.eu.
The propagation paradigm
This project falls within the theoretical framework of propagations, as defined by Pr Boullier. This paradigm proposes a third path for the social sciences, alongside approaches based on structures (institutions, classes, norms) and individual preferences (rational choices, emotions, beliefs). It places at the heart of the analysis the circulating entities - technical objects, rumours, values, digital content - that pass through collectives and transform them.
Propagations are thought of as dynamic, viral or memetic phenomena, which can now be better tracked and analysed thanks to digital traces. They allow us to understand social life in terms of flows, life cycles and shifts, and not just in terms of stable structures or isolated players. It is also an invitation to combine social sciences with data sciences, theories of imitation and the digital infrastructures that make these phenomena observable.
Project objectives
Pr Boullier wants to test methods for detecting propagation patterns, by exploiting the APIs of applications such as Reddit, 4chan or Mastodon, where the effects of virality can be observed and documented. Initially, members will not be analysing content directly, but rather aggregating traces and experimenting with modelling tools (machine learning, lifecycle analysis, inferences based on tipping points, etc.).
A second phase of the project will mobilise linguistic engineering methods (not necessarily LLM) to track the mutations of topics or memes as they are propagated, without having to reconstruct the networks of actors.
A design to combat viral reactivity
The designers will be tasked with thinking up a mobile dashboard designed not to accelerate virality, but to slow down reactivity, offering a different way of interacting with feeds. The challenge is to design critical interfaces capable of visualising propagation cycles, while at the same time offering a form of de-acceleration of usage.
Let me know if you want to join via videoconference, it might be interesting