The voice of stereotype in digital assistants

Abstract

In digital assistance devices, human voice reproduction and synthesis processes do not just are ruled by purely technological mechanisms. They also show both ways of conceiving language according to information patterns and ways of updating and perpetuating certain social stereotypes, particularly those of gender. It is with the aim of inquiring into the articulation of both that the contents exposed in this paper are justified, as well as the main conceptual distinction, which is developed through it, between ‘the voice that hears’ and ‘the voice that is heard’. Two significant theses guide such purpose. The first suggests that the standardization of digital assistants according to gender discrimination criteria is still, in our times, one of several expressions of the broad phenomenon of the sexual division of labour. The second thesis, in turn, covers the technological phenomenon of the so-called “digital convergence” and concerns how auditory stereotypes, based on the image of the mother-caregiver, complement and reinforce visual stereotypes. For the theoretical articulation of these two theses to be feasible, it is necessary, above all, to demystify the supposed verbal interaction between machine and user, showing – contrary to what is inferred from the contemporary technological imaginary – the dialogical impossibility of both.