Siren Songs and Echo’s Response: Towards a Media Theory of the Voice in the Light of Speech Synthesis

Abstract

In contrast to phonographical recording, storage, and reproduction of the voice, most media theories, especially prominent media theories of the human voice, neglected the aspect of synthesizing human-like voices by non-human means. This paper takes this lacuna as a starting point for an inquiry into the media theory of (non)human voices under the premise that the epistemological difference between techn(olog)ical voice production and its mere re-production is illuminated by the mythological motifs of the Sirens and Echo, respectively. Interestingly, the interconnection between terror and tempting nonhuman voices, which is implemented in the cultural imaginary through the Sirens’ song, can be identified in the media history of speech synthesis, which challenges the idea(l) of the human voice as an anthropological constant. The main concerns here are to re-read the critique of Derrida’s Of Grammatology and other theories of the human voice in the light of speech synthesis and show how the oft-used term ‘disembodied voice’ is inadequate when it comes to describing phonographical, radiophonic, and telephonic hearing situations.

Keywords

• speech synthesis
• phonography
• voice theory
• embodied voices
• speaking machines
• technotraumatic affects