Socio-Semantic Network Analysis of Meaning Structures

Intervention de Nikita Basov, résident 2023-2024 de l'IEA de Paris, dans le cadre du séminaire "Systèmes complexes en sciences sociales" de l'EHESS.

Abstract

For centuries, social structure and culture are theorized as intertwined. Empirical analyses of the two, however, have been disjoint. While analyses of social structure relied heavily on quantitative methods, studies dealing with culture remained predominantly qualitative. The change, becoming prominent with the outburst of big data analysis and digital humanities, started even earlier. On the one hand, semantic network scholarship was evolving in computational science and linguistics. On the other hand, sociology put forward a structural view on culture as an ensemble of relations rather than a mere collection of elements such as concepts or words. These developments inspired quantitative approaches to culture, treating it as a structure in its own right, dual to social structure. This, in turn, opened a path to explanatory analyses of the interplay between social and cultural structures. Notwithstanding, social and cultural structural approaches remain loosely connected. Socio-semantic network analysis explicitly puts the two structures together. It is based on the idea of their ontological duality: Social relationships constitute the meaning of semantic associations with regard to the social context in which these relationships are embedded (e.g., ‘friendship’ implies co-presence in a school class but not online), while semantic associations shape the meaning of social relationships (e.g., ‘we are friends’) against the background of other semantic associations (e.g., ‘friend in need is a friend indeed’). Therefore, any account of meaning (and therefore, understanding of society) is to combine both social and cultural components, otherwise depriving either of meaning. This is particularly the case for structural analysis as it already deals with a simplified representation of reality, in contrast to ethnography. Meanwhile, inductive search for statistical regularities in structure is essential to capture implicit mechanisms that mould meaning and to put competitive theories to joint testing. This work, however, is not to exclude in-depth qualitative insight into context – to enable interpretive understanding. The talk will showcase how mixed socio-semantic network analysis unveils meaning and discuss some of the prospects for future development.

Programme détaillé et inscriptions