How to problematize a border, or: some ongoing notes on transient methodologies

Abstract

In this paper I rehearse a method for studying media that is engaged by what I am calling a transient methodology. This work with “transients“ emerges as much as an analytical tool as it does as a material engagement, in direct connection with my encounter with a GDR-era synthesizer prototype, the Subharchord I/II, and its use of a specific audio filter design which is found again only forty years later as a building block for most state-of-the-art speech recognition algorithms. I am particularly interested in one such algorithm, namely the so-called “Dialect Recognition Software”, a proprietary machine learning solution in use since 2017 by the German Office for Migration and Refugees in cases of undocumented asylum seekers in the country. By bringing these two stories closer and mobilizing them through the work of transients – a concept that works in sound as much as it does in movement – I seek to activate an experimental, speculative media archaeology that focuses on the haptics of listening. Transient methodologies, I argue, offer the potential to critically listen to these two technologies together while at the same time slowing them down for critical inquiry and analysis.

Keywords

• machine listening, sound studies, migration studies, media archaeology