The proof of enunciation. Digital photography, deepfake, and the refusing or renegotiation of the document

Abstract

In this paper we resume the arguments we made in a past research on the ambiguous epistemological status of digital photos, focusing on two main practices in which the enunciative device shows a manipulation moving from previous visual instances referentially grounded: mash-up and deepfake.
In the first case, an ironical utterer performs a sort of explicit fiction: he then denies any documental status to its photographic text, and in so doing obtains a pragmatic effect of ludic nature (i.e. funniness, or irony).
Deepfake, on the contrary, is a sophisticated form of visual simulation where there is no personal, intentional utterer: rather, it results from a hybrid enunciative framework, in which competences and performances are distributed among heterogeneous actants, through delegations and mediations.
Deepfake indeed is a sort of objective camouflage: it simulates the testimonial value of its referential content, hiding the complete unreliability of it and obtaining a pragmatic effect of lie (‘seeming’ + ‘not being’).