Voice User Interfaces (VUI) are slowly emerging into today’s IT applications. A decade ago, members of the speech science community (as well as market analysts) predicted a much faster growth of commercial VUI usage and are now wondering why the acceptance was so moderate during the last five years. One reason for this is certainly the difficult economic situation since the crash of the new technology market. Another reason is the fact that people are slow, especially in the light of an uncertain economy, to leave well-trodden paths for new shores. In this paper we raise the unorthodox question whether VUIs might work more efficiently, be better accepted by users and have a greater commercial impact on future IT markets if there existed a widely agreed communication paradigm on how to use VUIs. For the moment we’d like to call this paradigm ‘Lingua Machinae’.