Dumb intelligence? Translation as technological mediation

Abstract

We propose a semiotic approach to understanding and assessing language technologies. Specifically, by adopting a recent semiotic and broad concept of translation, developed by Kobus Marais, we bring semiotic theory into the service of philosophy of technology. Our perspective reveals that commonly assumed expectations about language generative technologies are mistaken and misleading when shaped through an ideal of engineering humanlike interlocutors, which we illustrate with examples. We find that (software) engineering pursues this ideal, which, fuelled by classical humanism, assumes that language is an anthropic marker. By explaining (technological) emergence as a semiosic process, we develop a robust underpinning for the Mind–Technology Thesis, namely refuting mind-and-matter substance dualism through an evolutionist perspective that construes technology as mind-work. In this vein, semiotics corroborates with externalist theories of mind and postphenomenology in understanding mind and technology as mutually intrinsic. This leads to a semiotics-grounded advocacy of the view in philosophy of technology, championed by Elena Esposito, that for artefacts properly to communicate with biological organisms, they do not require “intelligence”.