Urban highways are common, especially in the US, making cities more car-centric. They promise the annihilation of distance but obstruct pedestrian mobility, thus playing a key role in limiting social interactions locally. Although this limiting role is widely acknowledged in urban studies, the quantitative relationship between urban highways and social ties is barely tested. Here we define a Barrier Score that relates massive, geolocated online social network data to highways in the 50 largest US cities and which captures the connectivity difference between the actual network and a randomized null model. As such, at the unprecedented granularity of individual social ties, we show that urban highways are associated with decreased social connectivity. This barrier effect iis especially strong for short distances and consistent with historical cases of highways that were built to purposefully disrupt or isolate Black neighborhoods. Using multivariate regression models at the census tract level, we confirm that decreased social connectivity related to the presence of highways holds even when accounting for confounding factors such as socio-demographic backgrounds, or other types of barriers. By combining spatial infrastructure with social tie data, our method adds a new dimension to demographic studies of social segregation. Our work can inform reparative planning for an evidence-based reduction of spatial inequality, and more generally, support a better integration of the social fabric in urban planning.

Although neighborhood relations are omnipresent in debates on the intensification of urban segregation in cities, few studies compare them to other types of social relationships and circles (family, friends, professional, associative, etc.), whether they are locally based and/or extend to other areas. However, as Barry Wellman pointed out as early as 1979, taking social networks as a starting point allows us to connect relationships shaped by residential contexts to the overall structure of personal networks analyze their role in network formation, and assess their impact on access to various resources.

The research underlying this presentation studies the spatial dimension of social segregation, using the conceptual and methodological approaches of social network analysis. More specifically, it compares relational contexts with local activity contexts to examine the place and role of neighbourhood relationships in networks to see whether and to what extent these relationships participate in the formation of "entre-soi" (understood as a sociability that is both homogeneous and cohesive), which primarily takes place within shared activities (education, work, etc.). The aim is to expand on the results of a previous questionnaire survey in which we were able to distinguish between two types of local relationships: relationships created with "neighbours" and those formed with what we call "more-than-neighbours", and those beyond the local space. It has also highlighted the importance of activity contexts, - beyond mere geographical proximity (work, friend groups, associations, etc.) - in reinforcing the homogeneity and cohesion of personal networks. However, quantitative data did not allow us to study precisely each of these local relational contexts and how they interact. It also did not distinguish between formal and informal sociability, nor between individual and collective relations.

Based on an ongoing qualitative study conducted in a neighborhood in Toulouse (France), this presentation proposes a more detailed and dynamic analysis of the plurality of activity contexts in which local relationships are embedded, as well as their interconnections. The goal is to understand how these relationships are formed, maintained, and evolve over time, and to what extent they contribute to either social homogeneity or openness. Interviews with residents and discussions based on the visualization of their network have enabled us to draw up sociological portraits revealing different relational dynamics depending on the type of relationship, social circles and also ego characteristics. These first results provide new insights into the interplay between "structural" homogeneity (linked to the neighborhood's social composition) and "elective" homogeneity (driven by affinity-based ties) within personal networks.

The academic research ecosystem has long been shaped by systemic inequities, favoring scholars with access to well-funded institutions, elite networks, and dominant publication platforms. As a result, knowledge production and dissemination remain concentrated among a privileged few, reinforcing epistemic hierarchies and limiting diverse contributions. In an era of geopolitical shifts and the rapid advancement of Generative AI (GenAI), new opportunities emerge to rethink how academic collaboration is structured.

This study suggests to employ a mixed-method approach – combining qualitative insights from interviews with a quantitative social network analysis to identify key structural barriers faced by researchers from underfunded institutions. By mapping these obstacles, we aim to understand how digital inequalities, funding constraints, and AI-driven changes affect academic participation. Based on these findings, we propose BrainSpill, a prototype social network platform designed to foster inclusive, decentralized knowledge exchange. This tool seeks to mitigate existing disparities by leveraging AI-driven matchmaking, open-access collaboration tools, and reputation mechanisms that promote fair participation.

Our work contributes to the discourse on equity in academia by offering actionable solutions to bridge the gap between privileged and underrepresented researchers, ultimately advocating for a more inclusive and globally connected research landscape.

Recent scholarship has increasingly examined global disparities in academic production between central and peripheral regions. Empirical studies highlight inequalities in research impact, Anglophone publishing bias, linguistic barriers, and North-South divides in the generalizability of knowledge claims. On the theoretical and political front, “decolonizing knowledge” has gained traction across disciplines, including sociology, as scholars interrogate the institutional and epistemological underpinnings of these disparities. This study applies insights and tools from social network analysis to examine between-country collaboration networks in sociological publishing. As a first step, I analyze co-authorship patterns in two US-branded but globally prestigious journals—American Sociological Review (ASR) and American Journal of Sociology (AJS) (1966-2018), assessing the geographic distribution of authors and their institutional affiliations. Findings reveal that the U.S. is an absolute star in this network, with nearly all international collaborations involving U.S.-based institutions. Only three countries have ever published independently in ASR/AJS without U.S.-affiliated co-authors, while Austria is the only case of international collaboration without a U.S. tie. Frequent U.S.-linked collaborators include developed countries like Canada, the UK, Germany, and Israel, while Global South countries are largely absent. European nations form the only visible non-U.S. collaboration cluster. As a next step, I will apply Borgatti’s core/periphery model to quantify network centrality of each participating country and extend the analysis to British Journal of Sociology (BJS), Sociology, Journal of Sociology (JoS), Current Sociology (CS), and International Sociology (IS) to assess whether these non-U.S. flagship journals exhibit more globally inclusive patterns.