This paper reports on a multi-faceted study of enumerations involving linguists, psycholinguists and computer scientists. Our first point is that layout must be seen as a combination of lexicosyntactic and visual features, which we call "formatting", rather than restricted to visual features. The article lays down the theoretical bases for a model of text architecture making explicit the relations between discursive and visual formulations. The corpus-based study of enumerations, which considers standard and non-standard forms, enables us to put the architecture model to the test. We identify major markers (typographical, layout, lexico-syntactic markers), and arrive at fine-grained characterization and a classification of enumerations. As regards text organization, we show that in order to arrive at a complete and precise representation, we need to articulate and integrate the two models used: the architecture model and RST.