The algorithm is a text

Abstract

In this essay, I argue that algorithms can and should be read as performative texts, whose discursive actions organize meaning, shape interaction, and impact social practices. Drawing on a contemporary framework of Text Linguistics, which understands text as a situated, intersubjective, cognitive, and multimodal activity (Koch, 2006; Marcuschi, 2008; Beaugrande, 1997), I maintain that algorithms are not merely coded instructions, but forms of textualization that exert symbolic and pragmatic power in the digital world. To support this hypothesis, I analyze two algorithmic simulations written in programming language: a coffee-making model and a content recommendation system for streaming platforms. Through these analyses, I demonstrate how textuality factors such as cohesion, coherence, intentionality, acceptability, informativeness, situationality, and intertextuality emerge as effects of communication processes mediated by code. By shifting the focus from structural criteria to meaning-making dynamics, I show that algorithms not only share textual features with human discourse, but also generate interpretations, responses, and experiences. I conclude that recognizing algorithms as texts enables us to submit them to critical and ethical scrutiny, capable of revealing embedded ideologies and demanding transparency and the rewriting of the codes that shape contemporary digital life.