The present contribution starts from the general observations (i) that the study of text varieties has commonly emphasized the mode (speech vs. writing) as an essential variable and (ii) that linguistic analyses increasingly consider performed language (i.e., fictional scripted material as represented in telecinematic language and lyrics, for instance) as an object worth studying in its own right. It is recognized (i) that assessing performed language in terms of the traditional spoken-written dichotomy fails due to a number of inherent properties of relevant text varieties and their circumstances of production and reception, and (ii) that schemes applied within conventional (bottom-up) register perspectives may not be fully adequate for a text-linguistic approach toward performed language either. Koch and Oesterreicher's communicative model (KOM), which takes account of contextual factors as well as of specific linguistic strategies and establishes a continuum between the language of distance and the language of immediacy, as well as three attempts at modifying KOM are introduced. It is argued that both KOM (as a top-down model) and its modifications have weaknesses when dealing with performed text varieties. An enhanced communicative model, including the dimensions synchronicity and authenticity, is sketched out, which could serve both as a starting point and complement for empirical investigations of textual variation whenever performed language is included.