Italian has recently witnessed a steady increase in the use of unfamiliar morphological blends from names and nouns. While they serve an identificatory and descriptive function (in the sense of Anderson 2007), blends are created in extragrammatical morphology with careful attention to the semantic concepts encoded by the individual SWs, understanding blends from names and nouns depends on the decoder’s direct or surrogate experience of the related reference. Significantly, blends are coined out of the need to be relevant (Sperber/Wilson 1990) and show various degrees of morphotactic and morphosemantic transparency (Thornton 1986; Dressler 1987, 1999), which makes them memorable (Lehrer 2003). In this paper we therefore address blends from names and nouns within the framework of the Naturalness Theory (Thornton 1986; Dressler et al. 1987; Dressler 1999). As will be seen, although blends are not created in rule-based grammars, some overall preferences and regularities can be observed for more core items (see Bat-El/Cohen, in press, within the framework of Optimality Theory) under the principle of saliency (Dressler 1987). Focusing on their morphosyntactic transparency, we provide a typology of Italian unfamiliar blends from names and nouns functioning as naming units in order to delimit the category and reassess current typologies. While we allow for a continuum of morphotactic transparency within the prototypical category of extragrammatcial subtractive word-formations, we slightly adapt Ronneberger-Sibold (2006) and suggest we distinguish between blends on the one hand and the neighbouring category of clipped compounds on the other. Moving on to morphosemantic transparency and conceptual motivation, we use concepts from Cognitive Grammar, Cognitive Metaphor Theory and theories of Conceptual Blending (cf. Lakoff/Johnson 1980; Langacker 1987; Ruiz de Mendoza 1998, 2000; Kemmer 2003) to provide some preliminary remarks on specific subtypes, and blends from personal names used in journalese and media language in particular (e.g. Berluscotti < Berlusconi + Bertinotti, Berlingotti < Berlinguer + Bertinotti). The data suggests that we cannot as yet talk about instantiations or extensions of entrenched schemas