—A Case Study on Modelling Language as a Complex Adaptive System
The story of Artificial Intelligence is often told as a march of breakthroughs and inevitable progress. But beneath this story lies a different picture: AI and other sciences are better understood as a living web of research threads, where ideas emerge, recombine, and sometimes fade from view. Understanding how techno-scientific research fields function requires novel conceptual tools, such as phylo-epistemic networks – graph-based structures for mapping out the interactions between research communities, conceptual networks and technical innovations.
In this talk, I will present a case study of such a phylo-epistemic network that traces the intellectual history of modelling language as a complex adaptive system – one of the central research projects that has been developed at the Sony Computer Science Laboratories, Paris Lab, since its founding in 1996, and the VUB Artificial Intelligence Laboratory in Brussels. Starting from a simple but transformative question - how can a population of language users self-organize a language from scratch - this case study demonstrates how shifts in perspective and novel discoveries can lead a research stream in a radically different direction than the mainstream.
By understanding the history of the science of AI as a living epistemic ecosystem, case studies such as these also direct our attention to the future. As today’s dominant paradigms crowd out alternatives, the field’s epistemic diversity grows fragile. Phylo-epistemic analysis makes this fragility concrete: it can identify which research lineages went extinct; often not because of failure but because of institutional starvation and lack of resources. This talk therefore makes a plea to look at our field’s history not as a closed book, but as a living resource to help us think differently about what AI is and what it can become.
About Remi van Trijp
Remi van Trijp is Research Leader at the Sony Computer Science Laboratories in Paris. His work is situated at the intersection of computational linguistics, cognitive science, and artificial intelligence, with a focus on construction grammar, language evolution, and computational models of meaning-making. He develops multi-agent systems that treat language as a complex adaptive system, combining linguistic theory with experimental, evolutionary, and emergentist-symbolic approaches to language understanding. He has participated in several European research projects (including ECAgents, ALEAR, ESSENCE, and MUHAI) and has published extensively in peer-reviewed venues.
He is a co-author of the recently released open book “History of Ideas in the Science of AI” VUB Press, Zenodo link.
Details
Date and Time: Thursday, 19th of February 2026 - 11 AM
Location: Sony CSL, 6 rue Amyot, 75005 Paris
Registration: here