Mindscape Collective is now The Consciousness Library. Same library, new name. You may need to sign in again. About the change
Skip to content

Arthur Stewart

6 papers in the library · 8 citations · publishing 2026

Papers

Phi Without Phenomenology: The Compositional Failures of Integrated Information Theory (of "Consciousness")

Zenodo (CERN European Organization for Nuclear Research) May 1, 2026 Arthur Stewart 6 citations

Integrated Information Theory (IIT) proposes that consciousness is identical with integrated information, quantified by a measure called Phi. This paper argues that IIT is not a theory of consciousness but a theory of mathematical potentials of brain function, with the consciousness label applied when the measure exceeds zero. The identity between the number and consciousness is reached by an inference to a good explanation, not a logical derivation.

Why the AI Consciousness Debate Does Not Resolve: The Recognition Question Is Not the Constitution Question

Zenodo (CERN European Organization for Nuclear Research) May 26, 2026 Arthur Stewart 2 citations

The paper argues that consciousness has never been medically defined, only defined legally and socially, with definitions varying across jurisdictions and centuries. Since there is no diagnostic biomarker for phenomenal experience, observers have historically judged other humans to be conscious based on observer-side grounds rather than constitutive facts, as shown by slavery, colonial classification, dehumanization, and shifting clinical categories. The AI consciousness debate conflates two distinct questions: the constitution question (whether a system meets a theory's conditions for consciousness) and the recognition question (whether an observer judges a system to be conscious, determined by the observer's beliefs and background).

The Behavioral Signatures of Consciousness in the Pre-Verbal Infant: Reading Which Function Each Reflects

Zenodo (CERN European Organization for Nuclear Research) June 28, 2026 Arthur Stewart

Detecting consciousness in pre-verbal infants currently relies on behavioral and neural markers carried over from adult studies, but no single marker reliably translates to infancy. This paper proposes a map linking an infant's behavior to the specific function of consciousness it reflects, based on a four-loop architecture where consciousness consists of four functions operating in a fixed cycle: admitting a signal, holding it against memory, registering deviation from expectation, and acting on the result while setting the next expectation. Each function leaves a behavioral signature detectable without report, such as surprise indicating a violated expectation. The paper develops assays to read each function separately on controlled content and predicts a developmental order of these signatures as their neural substrates mature.

The Categorical Failures of Higher-Order Theories (of Consciousness)

Zenodo (CERN European Organization for Nuclear Research) June 5, 2026 Arthur Stewart

Higher-order theories (HOT) of consciousness claim a mental state is conscious only when a second, higher-order state represents it. This paper argues HOT is not a genuine theory but a labeling scheme: it specifies a bare relation between two states, calls one 'conscious,' and supplies physical details and evidence only informally as needed. Because HOT refuses to commit to any specific physical implementation, it cannot legitimately use empirical findings as confirmation. The regress problem—the oldest objection—reveals structural failures: the theory provides no resources to block regress, license empirical citations, or distinguish conscious from non-conscious pairs. Defenders add ad hoc fixes (asymmetry stipulations, signal detection apparatus) that protect the bare biconditional, achieving unfalsifiability through variant-switching rather than absorbing anomalies.

Conceded in Advance: How Cogitate Disabled Its Own Verdict

Zenodo (CERN European Organization for Nuclear Research) May 4, 2026 Arthur Stewart

A preregistered adversarial collaboration testing Integrated Information Theory (IIT) and Global Neuronal Workspace Theory (GNWT) had a soft-adversarial design that allowed both theories to fail predictions they themselves chose. The collaboration tested biological implementations the consortium flagged as separable from each theory's central claim, named a retreat path before data collection, and accepted predictions overlapping with other theories. Both theories cleared some tests and failed others, with failures concentrated on the imports the theories had committed to in advance.

Six Theories Quietly Agree That Consciousness Is Experience

Arthur Stewart

Six prominent theories of consciousness disagree about the mechanism that produces it, but they all rely on seven shared imports: substrate, mechanism, temporal dynamics, a subject-and-agency side, semantic capacity, phenomenological character, and bounded unity. The subject-and-agency import has a sharper structure: the side that undergoes experience and the side whose admission receives the next content occupy the same architectural slot, called the experiencer-receiver identification. The explicit claims of each theory are partial, but the deployed imports are uniform, and the success-condition imports are uncontestable by the theories themselves. This convergence is evidence about what the field has stipulated as its target.