Prosodia

A “synthetic” performer named Prosody, two actresses in the flesh, a theatre director, and a traditional musician set out to stage an ancient epic: The Epic of Siri – a vast narrative celebrating the deeds of a woman, still sung today in South India and rivalling the Iliad in both scale and antiquity. As they attempt to synchronise and sing together, the characters – human and non-human alike – begin to reflect on the technologies that underpin their practices, namely three generations of performative artificial language: “sung-narration,” “acting,” and “synthetic speech.” Prosody’s inability to cry becomes a technical and artistic obstacle, since every epic traditionally ends in tears. The ensemble eventually discovers a solution in one of her oldest inherent traits.

In Prosody, primordial patterns of speech resurface within the synthetic voices generated today through artificial intelligence, echoing the rhythmic and formal patterning of human storytelling – from ancient epic poetry to contemporary influencer content. In the process, the work also dispels many of the myths surrounding AI, particularly those propagated by the profit-driven entities that develop it.

For the purposes of the piece, visual artist Nicoline van Harskamp has developed a stage mechanism in which the lines of one character – a list of their turns in a dialogue – are connected to a text-to-speech system. At the press of a wireless button, the actress triggers an offstage computer, initiating the live generation of a line delivered by their synthetic counterpart. As with a human performer, the prosodic qualities of each utterance – its volume, pitch, intonation, and tempo – remain variable and unpredictable, even though the textual content itself is pre-scripted.

When operating this system live, the actress effectively shapes the delivery of their non-human counterpart while performing their own lines. What emerges is a peculiar form of monologue unfolding as dialogue, requiring acute timing and a high degree of self-direction. With recent advances in AI technologies, the stage mechanism has been upgraded to enable direct speech-to-speech transformation. In this way, Prosody can modulate her vocal expression in response to the prosodic cues she detects in the voices of the human performers on stage.

In Prosody, technology is neither condemned nor glorified. Instead, it is deployed with precision and clarity to invite a reflection on how we speak, how we listen, and how – ultimately – our voices are shaped by the media that carry them.