The vast majority of the discourses of today's mediatized world reach us through screens and frames: screens are the devices of circulation and reception of almost all the media encompassing the three communication systems that make up contemporary mediatization as described by the Argentine scholar Mario Carlon: underground, social media and traditional mass media. Taking as starting point this increasingly ‘becoming-screen’ of our present mediatized environment, we observe with Carlon (2020) that the enunciative materiality of screens is what, in our current digital culture, confers visibility and 'autonomy' to the ‘media circulation of meaning'. On the basis of this inseparability between such screen media proliferation and our deeply mediatized worlds, we propose in this paper to explore the emergence and trajectory of the topological and relational concept of 'non-anthropoid enunciation'. By following the historical legacy of semioticians such as Metz, Veron and Traversa, we seek to understand the complex processes of enunciation in today's 'hypermediatized' societies. The underlying hypothesis is that the historical meta-process of mediatization ('scale ruptures') is rendering evident that enunciation is not simply a deictic expression of a prior identity or subject, but a 'meta-discursive' activity present in any discourse, which only other discourses can set in motion (in recognition/reading/reception). With examples taken from a filmic corpus, I will try to illustrate that enunciation are these enfoldings of (audio-visual) utterances back on themselves, without any personological guarantees, since any enunciation whose 'source' ('foyer', Metz) is a screen renders visible the non-anthropoid character of enunciation tout court.