A series of studies converge in the finding that, during the Covid-19 emergency, compliance with the mandate to wear a face mask generally leads to lower compliance with the recommendation to keep a minimal physical distance from others. In the present state of knowledge, we know that this form of risk compensation occurs but we lack a proper understanding of why it occurs. GeRICO’s scientific aim is to explain the distance-reducing effect of the face mask. From a public policy perspective, GeRICO’s applied aim is to issue recommendations to attenuate the risk-compensatory effects of the co-existence of a face mask mandate with the one-meter rule. Bringing together concepts and methods from sociology, psychology, political science and linguistics, the proposal puts to test two candidate mechanisms: 1) people may approach more when the mask is worn because of a “false sense of security” (operationalized as a risk perception); 2) people may approach more in the presence of the covering because the face mask, by attenuating the voice, degrades speech intelligibility. Further, the project investigates whether gender, social status and perceived ethnicity moderate the distance-reducing effects of the mask. Methodologically, GERICO relies on an original and complementary combination of field, online and virtual-reality experiments.

The OriKunda project aims to revise the unique history of the Kunda people and the Kunda language from their genesis to the present day, through historical linguistics, anthropologic linguistics and sociolinguistics. The Kunda were troops of slave soldiers originating from different tribes and serving in the territories occupied by Portuguese settlers (‘prazos’) in central Mozambique during colonial times. Out of their common social identity was born an ethnic identity involving the creation of a vehicular language, Kunda, resulting from an intra-Bantu interbreeding. With the collapse of the ‘prazos’ system in the 19th century and the liberation of slaves, the Kunda retreated westward at the confluence between the Zambezi and Luangwa rivers, which today corresponds to the cross-border area between Zambia, Mozambique and Zimbabwe. The OriKunda project proposes to put linguistics at the service of history by redrawing the kunda epic through their language and their testimonies. What is Kunda? What does this hybrid language, created less than 4 centuries ago from typologically similar and genetically related languages, look like? What languages have participated in its development? Where can it be located in the genealogical classification of Bantu languages? What is the level of vitality of Kunda today, on the Zambian side, the Mozambican side and the Zimbabwean side? What presence, practices and perceptions of the language can be observed? What founding myth and what oral tradition do the Kunda convey? So many fascinating questions to which the OriKunda project intends to answer.

The SYSPOE project takes place in the renewal of police studies in the field of social sciences. It aims at studying police systems, defined as configurations composed by the various actors of policing in a given space, in Europe and its colonial possessions in the 18th and 19th centuries, in a comparative perspective and at the crossroads of different disciplines. Supported by 4 research units, this project gathers 14 permanent researchers, historians of the 18th and 19th centuries, but also a sociologist and a political scientist. It combines a general, interdisciplinary reflection on police systems through a research seminar, and specific archival research delimited by 5 thematic workpackages : police systems and circulations ; police systems and colonial territories ; plural policing ; military culture and police systems ; police systems, crisis, revolutions and disasters. It aims at building the preliminary fundations for a European history of police forces, contributing to a better comprehension of European societies in the 18th and 19th centuries by observing their forms of regulation, and illuminating by the expertise of historical reflection some issues on the contemporary police systems.

The ability to infer cause-effect relations is an important facet of learning. In particular, learning the causal effect of our behaviors provides the basis for rational decision-making and allows people to engage in meaningful life and social interactions. A number of psychological theories have been proposed to account for the internal representations mediating causal learning. However, an integrated understanding linking causal learning theories and brain systems is still lacking. The CausaL project will investigate the neural and computational bases of causal learning in the context of goal-directed behaviors (action-outcome causal relations). To do so, we will need to lift two key barriers. The first is the lack of neurocomputational models that formalise psychological theories into predictions about the brain computations. The second is the lack of clear understanding of the brain dynamics supporting action-outcome causal learning. Two preliminary studies of our group have set the ground for lifting these barriers and demonstrated the feasibility of the project. The CausaL project will pursue this work along two Tasks. In Task 1, we will confront two leading neurocomputational learning frameworks: Active Inference (AI) and Reinforcement Learning (RL). Both approaches can be used to formalise the relation between decision variables predicted by causal learning theories and learning behaviors in humans, but differ in the underlying conceptual structure. Put simply, whereas the first is a Bayesian approach aiming at the minimization of uncertainty during explorative and exploitative behavior, the latter formalizes learning as a process maximizing cumulative rewards. By means of simulations of ideal agents and comparison with behavioral data collected from a large cohort of participants (N=180), we will develop neurocomputational models of action-outcome causal learning. We will then compare AI and RL models in their ability to accurately explain empirical behavioral patterns, choice patterns and causal scores. Finally, we will exploit the “best” models to yield predictions about the underlying neural computations. In Task 2, we will test AI and RL models on brain data and study how learning-related brain regions interact. In fact, it is now acquainted that action-outcome causal learning is a brain network phenomenon. However, it is still unclear how fronto-striatal regions dynamically interact to support learning computations. We will investigate brain data to test whether the internal representations predicted by psychological theories and implemented in AI and RL models (for example, conditional probabilities between actions and outcomes, causal beliefs) are encoded in: i) functional connectivity dynamics and directional influences between learning-related brain regions, by means of magnetoencephalography studies in healthy participants and intracranial stereo-electroencephalography in epileptic patients; ii) distinct spatial patterns of brain activaty along fronto-striatal territories, by means of functional and diffusion magnetic resonance imaging (MRI). The combination of functional and structural brain data will reveal how causal learning emerges from interplay between large-scale functional interactions and anatomo-functional gradients along fronto-striatal circuits. To conclude, the CausaL project offers the extraordinary opportunity to link theoretical models of action-outcome causal learning, behavior and brain network dynamics, the so-called cognitive architectures of causal learning (CausaL).

The evolution of scientific knowledge is directly related to the history of humanity. Document archives and bibliographic sources like the “Web Of Science” or PubMed contain a valuable source for the analysis and reconstruction of this evolution. The proposed project takes as starting point the contributions of D. Chavalarias and J.P. Cointet about the analysis of the dynamicity of evolutive corpora and the automatic construction of “phylomemetic” topic lattices (as an analogy with genealogic trees of natural species). Currently existing tools are limited to the processing of medium sized collections and a non interactive usage. The expected project outcome is situated at the crossroad between Computer science and Social sciences. Our goal is to develop new highly performant tools for building phylomemetic maps of science by exploiting recent technologies for parallelizing tasks and algorithms on complex and voluminous data. These tools are conceived and validated in collaboration with experts in philosophy and history of science over large scientific archives.