Until now, most studies carried onto social or semantic networks have considered each of these networks independently. Our goal here is to bring a formal frame for studying both networks empirically as well as to point out stylized facts that would explain their reciprocal influence and the emergence of clusters of agents, which may also be regarded as ''cultural cliques''. We show how to apply the Galois lattice theory to the modeling of the coevolution of social and conceptual networks, and the characterization of cultural communities. Basing our approach on Barabasi-Albert's models, we however extend the usual preferential attachment probability in order to take into account the reciprocal influence of both networks, therefore introducing the notion of dual distance. In addition to providing a theoretic frame we draw here a program of empirical tests which should give root to a more analytical model and the consequent simulation and validation. In a broader view, adopting and actually implementing the paradigm of cultural epidemiology, we could proceed further with the study of knowledge diffusion and explain how the social network structure affects concept propagation and in return how concept propagation affects the social network.