From Distant to Dialogic Reading: Large Language Models and the Methodological Reconfiguration of Literary Criticism

Abstract

This review examines the conceptual architecture and methodological implications of the dialogic reading framework as proposed in "From Distant to Dialogic Reading: Large Language Models and the Methodological Reconfiguration of Literary Criticism." The framework advances a timely intervention in digital literary studies by proposing a shift from the extractive logic of distant reading to a dialogic mode in which large language models function not as instruments of pattern detection but as interlocutors in a recursive hermeneutic process. Situating the framework within the broader scholarly conversation surrounding AI and literary interpretation, this review evaluates its theoretical foundations, methodological rigor, and empirical grounding. The framework draws on Bakhtinian dialogism, Eco's theory of the open work, and the functionalist stance in AI criticism to provide a vocabulary for treating machine-generated interpretations as provocations rather than pronouncements. However, the review argues that while the framework's central thesis offers a productive reconceptualization of computational criticism's aims, its contribution remains primarily programmatic. The absence of controlled comparison, the underspecification of criteria for interpretive value, and the insufficient engagement with critical AI scholarship collectively limit its immediate applicability. The Barthes–Eco Simulation, while pedagogically suggestive, does not demonstrate that dialogic reading produces insights unavailable through conventional methods. The review identifies three requirements for operationalizing dialogic reading as a testable methodology: specifying the conditions under which LLM outputs contribute interpretive surplus rather than algorithmic positivity; developing criteria for evaluating human-machine interpretive collaboration; and conducting controlled comparisons against established methods on common corpora. The hermeneutic hypercycle model offers one promising template for such empirical work. The framework provides the conceptual scaffolding; the harder work of building and testing the structure remains.