Toward a semiotic pyramid: language studies, AI, and knowledge exchange economy

Abstract

Abstract This paper addresses the interrelation between AI and language studies in the (digital) humanities in the context of knowledge exchange economies. We propose a three-dimensional semiotic model (pyramid) that includes AI as an active agent in meaning-creation into the sign system of communication and test its validity on autocompletion and predictive texting. We, therefore, analyzed 360 million tweets on COVID-19 over a period of 2 years and focused on the time-dependent shift of language use in the positive and negative extremes correlating to significant external events. The data suggest that digital language use has significantly changed over the course of the pandemic. Feedback loops are created between AI, utterance, and external reality that fossilize preconceived notions into textual isoglosses by suggesting and completing text entries. We thus argue that AI subtly changes the perception of reality by influencing the choice of wording. To mitigate the impact of AI in meaning-creation, we highlight the structural precautions undertaken by providers; we argue, however, that the (digital) humanities may re-emphasize their intrinsic value within the economy of knowledge exchange according to the subject-specific insights they may provide for language use and AI.