Italian sociolinguistics has been marked by a diamesic dimension for almost three decades. The aim of this paper is to provide an overview of the stages which – starting from the suggestion put forward by A. A. Mioni in 1983 – have led scholars to include medial differences within Berruto’s 1987 architecture of contemporary Italian. Much of the interest generated by that phase of research is grounded in the notions of written and spoken language (at times only alluded to or theoretically presupposed as a point of reference), which have converged into the two extreme points of the diamesic continuum. The objections which were initially raised seem to be justified by the now prevailing tendency to consider the diamesic variation as a subtype of diaphasic variation. The historical perspective adopted in this paper provides the opportunity to assess the merits and limits of diamesia, to explain why it is an all-Italian notion and to reflect on the self-same concept of variety.