Our research concerns the qualitative analysis of a corpus of spontaneous conversations, recorded in an urban environment in Sardinia (Italy). More precisely, our data were collected in Cagliari, the island's main city and administrative capital. Many elements are taken into consideration within the framework of our study, ranging from the specificities of the area of investigation concerned - the urban metropolitan area - to the theoretical and methodological questions which stem from this context of study, namely the multidisciplinary approach to this kind of research, the choice of the contexts of investigation, methodology of the data collection, etc. The Italian sociolinguistic situation provides an ideal viewpoint for the analysis of bilingualism phenomena, diglossia, language change, code-switching and codemixing, given the broad range of structural differences between the languages in contact. Our data consist of spontaneous conversations recorded in a variety of different interaction contexts. Moreover, the conversations may be ranked in various degrees of formality, having in one extremity, "formal" situations (public Institutions: employee - clients) and in the other, "informal" situations (school entrance/exit: students). We will compare our results obtained through an in-depth analysis of our spontaneous productions to the data stemming from other studies conducted in this area. In this way, we will present the evolution of the "Sardinian-Italian" linguistic repertoire and provide an overview of Sardinian as an "endangered" minority language, from a dynamic point of view, thus rejecting any sort of "archaeological" approaches. Rather, we aim to highlight Sardinian's presence in daily interactions and its close contact with Italian in the repertoires of speakers from different age-groups.