MOSAIC (Multi-Platform Observational & Sociodynamic Analysis of Individual Configurations) is a sociological framework for analyzing behavior across digital platforms. Rather than aggregating profiles or inferring psychological traits, MOSAIC operates at the positional level: its unit of analysis is not the individual as a stable subject, but the individual as a configuration of activated dispositions within structured socio-technical fields. The framework is developed through a dialogue with Lahire’s dispositional pluralism, Goffman’s dramaturgical sociology, Ricoeur’s narrative identity, Bourdieu’s field theory, Coleman’s macro-micro schema, and van Dijck and Poell’s platform logic. It introduces four core concepts - asymmetric positional cascade, illusory backstage, positional type-profile, and dispositional variance as analytical signal - and formalizes behavioral variance as: Vd (i) = Σᵢdem + Δᵢpse (Cₚ) + εbs The central thesis is that digital traces should not be interpreted as stable behavioral signals, but as structured positional variance across socio-technical environments. This paper presents the theoretical architecture of MOSAIC and includes an operational illustration based on anonymized cross-platform behavioral traces.
F. Forterre (Tue,) studied this question.