Authorship as Entanglement
The act of attributing a book to a named author — placing a single name on a cover, assigning copyright to a specific person, crediting a particular mind with the ideas the text contains — is, in Barad's framework, an agential cut performed after the fact on a process that was, during its unfolding, entangled. The cut is real and necessary: someone must be accountable for the claims a book makes, the errors it contains, the effects it produces. But the cut is also a practice, not a discovery, and it conceals as much as it reveals. In the AI age, when language models participate as active contributors to the process of composition, the concealment becomes visible in ways the previous authorship regime could keep hidden.
In the AI Story
The concept of the individual author, as McGann, Foucault, and others have documented, is historically specific — a relatively recent construction tied to copyright regimes, the commodification of texts, and the Romantic cult of individual genius. Medieval scribes did not claim authorship of the texts they copied and modified. Ancient texts were attributed to traditions, institutions, or gods. The author as sole, identifiable origin of a work is a practice that particular societies developed under particular conditions, not a universal feature of textual production.
Barad's framework locates this practice precisely: authorship is an agential cut performed by specific material-discursive configurations (copyright law, publishing contracts, academic credit systems) that produce the author as a determinate entity with determinate properties (responsibility, creativity, identifiable voice). The cut is performed on a process that was entangled — the author's words were shaped by editors, earlier readers, cultural context, institutional demands, technological affordances — but the cut produces the appearance of sovereign origination.
When AI enters the process as an active participant, the entanglement becomes visible in ways the previous apparatus concealed. Edo Segal's honesty about the collaboration with Claude — the moments when insights emerged that he could not cleanly attribute to either party, the laparoscopic surgery connection that felt like it belonged to the space between us — names what Barad's framework specifies: these moments are phenomena produced through intra-action, and the authorship cut imposes a boundary that the process did not contain.
The ethical stakes of this visibility are significant. If authorship is a cut rather than a discovery, then the cut could be made differently — attribution could be distributed across the apparatus, acknowledgments could track entanglement rather than perform sovereign origination, copyright regimes could recognize the constitutive role of contributors (editors, designers, the machine's trainers) that current frameworks render invisible. These are not utopian proposals but recognitions that the current cut is one option among others, sustained by specific institutional arrangements that could be reconfigured.
Origin
The framework draws on Barad's general theory of agential cuts (2007) and connects to established critiques of authorship in Foucault ('What Is an Author?', 1969), Roland Barthes ('The Death of the Author', 1968), and Jerome McGann's textual scholarship. The specific application to AI-assisted writing is contemporary, emerging in response to the entanglement that large language models make unavoidable.
Key Ideas
Authorship is a cut. It is performed on an entangled process, not discovered as a feature of it.
The cut is necessary. Institutional accountability requires naming someone answerable for the work.
The cut conceals. The collaborators — editors, cultural context, machine contributors — whose participation constituted the work are rendered invisible.
AI makes the entanglement visible. The previous apparatus could keep the collaboration hidden; AI's active participation cannot be concealed.
Different cuts are possible. The authorship regime is one option; others could be institutionalized with different distributions of credit and responsibility.
Appears in the Orange Pill Cycle
Further reading
- Karen Barad, Meeting the Universe Halfway (Duke, 2007)
- Michel Foucault, 'What Is an Author?' (1969)
- Jerome McGann, A Critique of Modern Textual Criticism (Chicago, 1983)