Operationalizing Consciousness: The Grounded Narrative Framework
Zenodo (CERN European Organization for Nuclear Research) June 12, 2026 DOI: 10.5281/zenodo.20672379 (opens in new tab) via OpenAlex
Summary
AI-generated from the abstractA new framework called the Grounded Narrative Framework proposes a physicalist account of consciousness that replaces unfalsifiable explanatory-gap arguments with three empirically testable diagnostic criteria. Unlike existing theories such as the Free Energy Principle, Global Workspace Theory, and Integrated Information Theory, this framework addresses the developmental conditions under which conscious processing arises, arguing that grounding is inherently bottom-up and cannot be engineered from the outside in. The framework offers a critique of current large language model architectures and provides a substrate-independent diagnostic approach applicable to any candidate system, biological or artificial.
Study at a glance
| Characteristics | Theoretical or philosophical paper Peer reviewed |
|---|---|
| Keywords | Operationalization Grounded theory Consciousness Narrative network Process computing |
| Key finding | Proposes the Grounded Narrative Framework as a physicalist account of consciousness with three empirically tractable diagnostic criteria, arguing that grounding is inherently bottom-up and cannot be engineered from the outside in. |
Abstract
This paper contributes to the literature by proposing the Grounded Narrative Framework, a physicalist account of consciousness that replaces unfalsifiable explanatory-gap arguments with three empirically tractable diagnostic criteria. Unlike existing frameworks — including the Free Energy Principle, Global Workspace Theory, and Integrated Information Theory — the Grounded Narrative Framework explicitly addresses the developmental conditions under which conscious processing arises, arguing that grounding is inherently bottom-up and cannot be engineered or instantiated from the outside in. In doing so, it offers both a principled critique of current large language model architectures and a substrate-independent diagnostic framework applicable to any candidate system, biological or artificial.