Wordless, A Conversation with Samuel Beckett and Morton Feldman at the End of History

Years of realization: 2018-2019

Commissioned by Asko-Schönberg Ensemble

Installation

Wordless … at the end of history is a sound installation and performance instrument featuring two text-to-speech (TTS) deep learning systems trained on the publicly available recordings of the voices of Samuel Beckett and Morton Feldman. A computer generates the dialogue for each speaker, and then remotely controls each of the tape machines to rewind and overdub the latest generated piece of dialogue on to the tape, which then is played back. The repeated overdubbings add up to an increasingly fragmented voice-like soundscape. The unintelligible Beckett and Feldman-like voice synthesizers enact an imaginary dialogue in a fictional data center at the end of human history, where the two endlessly recombine internet-scraped fragments of text relating to their work – including over 500 reviews of the two artists’ work, 300 Amazon.com user reviews, as well as publically released written letters by Samuel Beckett, and lecture transcripts by Morton Feldman. The dialogs are recorded, played back, and re-recorded on two robotically-controlled reel-to-reel tape recorders. The repeated overdubbings create an increasingly fragmented textural soundscape of human-sounding voice that evolves throughout the course of the installation.

Samuel Beckett, in particular, refused to be captured on tape or recorded in any form, thus making voice recordings nearly impossible to obtain save for a few short minutes. At the time of the installation’s creation, state-of-the-art text-to-speech and voice cloning using deep learning systems was in its infancy and required extremely large amounts of voice data in order to create a convincing replica, or even to converge to intelligible speech. Given an impossible task, recreating the voices of these two artists from intentionally limited voice data, the aesthetics and artefacts of the speech synthesis algorithms themselves become the center of the piece, producing eerie fragments of language, vocal cadence, timbres and resonances from two 20^th century artists who both had a complex relationship to voice, language and meaning.

Two Screen Installation

An alternative form of this installation was presented in Graz in 2021, presenting only the dialogue between the two artificial voice models on two monitors, emphasizing the two speakers and the aesthetics of the voice synthesis models.

Live Performance

In a small alteration to the installation, the computer-controlled reel-to-reel tape recorders become a live musical instrument. Here the installation becomes a performance instrument where the Beckett and Feldman text-to-speech synthesizers, as well as the tape machine overdubbing, can be performed in real time. In 2019, a number of short performances with the instrument were made together with Dutch spoken word poets Stokely Dichtman and Justin Samgar, building abstract vocal fragments and textures around their own poems reflecting on the meaning and absence of meaning in language and poetry.

The first edition of Wordless was comissioned by Asko-Schönberg Ensemble for the 2019 Words and Music biennale at the Muziekgebouw, Amsterdam