Thresholds of authorial originality in literary experiments with algorithmic writing

Abstract

Literary experiments with algorithmic writing, which characterise the contemporary digital discourse, descend from a long line of generative retroactions between reading and writing acts, whose distant origins date back to the medieval and the early modern period, with more direct origins in twentieth-century avant-garde movements of the 1920s and the beat generation of William Burroughs, for whom all writing was in fact cut-ups; a collage of words read, (over)heard. This collaborative sense of authorship has been especially appealing to the digital sensibility of recent decades, as more writers turn to computational tools to automatically recombine existing sources and generate probabilistic texts, with or without the subjective input of human composition. More recently, there has also been accelerated progress towards processing a type of Artificial Creative Intelligence, in which men and machines collaborate closely, disrupting romantic notions of authorship, legally anchored in the thresholds of originality. The present article addresses epistemic concerns arising from debates in the international literary scene about generative AI, integrating these discussions into a broader historical perspective on the ongoing conceptual transformations of authorship.