Methodology

A necessary-condition framework

Theories of consciousness are notoriously hard to compare because each tends to explain only the slice of the problem it is built for. This platform follows Yoshiyuki Ohmura's proposal: rather than scoring theories on a sliding scale, we ask whether each theory meets a fixed set of necessary conditions.

A theory is considered complete only if it can answer “Yes” to all 7 questions below. A “No” on any question does not make a theory bad—it marks a specific, well-defined gap that the theory needs to address. This is a deliberately strict, binary lens: each verdict is Yes or No, never a partial score.

Why not rank by testability?

Because consciousness is subjective, falsifiability is genuinely difficult for every theory of consciousness—and a theory can only be tested at all if it grants consciousness causal efficacy (question 3). For that reason we do not rank theories by testability. Among theories that satisfy the seven necessary conditions, the more meaningful comparison is the degree of unification and explanatory power—how well a theory coheres with adjacent sciences such as psychology and cognitive science. (That secondary comparison is not yet scored on the platform.)

The 7 Questions

All 11 theories are assessed against the same seven questions, each phrased so it can be answered Yes or No.

1. Phenomena: How can subjective experience be accounted for in physical or computational systems? (The 'hard problem': why physical or computational processes give rise to subjective experience at all.)
2. Self: Why does the subject of experience coincide with the initiator of action? (Why the 'I' that experiences is the same 'I' that acts — distinct from 'my body'.)
3. Causation: Does consciousness have causal efficacy within a system, beyond mere correlations or predictability? (Mental causation without epiphenomenalism; true causation distinguished from correlation and mere predictability.)
4. State: How can differences in levels or states of consciousness be explained? (e.g. wakefulness vs non-REM sleep or anesthesia, controlled by intrinsic mechanisms.)
5. Function: What functional or cognitive roles are associated with consciousness? (e.g. semantic integration, association, and cognitive control — and why they depend on consciousness in biological systems.)
6. Contents: How can the diversity, structure, and organization of conscious contents be explained? (e.g. the perceptual switches of binocular rivalry; integration, selection, and seriality of what reaches awareness.)
7. Universality: Can the theory be applied across different types of systems, including artificial systems? (A requirement on the form of explanation — applicable wherever the relevant causal and functional structures are present, not a commitment to substrate-independence.)

How a verdict is decided

Each verdict carries a short justification grounded in the theory's own source material. A theory earns a Yes only when its framework actually offers an account of the question—not merely a gesture toward one. When a theory's sources are silent on a question, or explicitly place it outside the theory's scope, the verdict is No.

Verdicts are produced by a disciplined, four-phase multi-agent pipeline running on Claude Opus 4.8. Every agent applies the same strict, source-grounded rubric, and the same per-question bar is enforced across all theories:

1. Dossiers — one agent per theory web-researches a factual dossier (core mechanism, what the theory actually says about each question, known limitations, and sources).
2. Judge — one agent per (theory × question) cell applies the strict rubric and that question's Yes-bar, returning a verdict with confidence, justification, key evidence, and what would flip it.
3. Calibrate — one auditor per question reviews all theories side by side and levels the bar, correcting any cell judged more strictly or loosely than its peers.
4. Verify — an independent skeptic tries to refute every verdict that calibration changed or that was marked low-confidence; the verdict then stays or flips.

The full, replicable specification—including the exact per-question bars, the structured outputs, the model, and how to reproduce a run—is on the verdict generation methodology page. After generation, every verdict can be challenged, refined, and overridden by the research community.

Fairness

• Identical prompts for each theory (only the theory name changes)
• Same seven questions asked of every theory
• Same rubric and the same binary Yes/No standard applied to all
• No theory-specific modifications to prompts or criteria
• Source-grounded: verdicts are justified from academic sources, not general knowledge

The 11 Evaluated Theories

Theory Selection

For details on how we decide which theories to include and exclude, see our Theory Selection Criteria page.

Important Caveats

• Completeness, not truth: answering all seven Yes means a theory is complete in scope—not that it is correct.
• A “No” is a gap, not a failure: it points to a specific question the theory has yet to address.
• Source-grounded: verdicts derive from academic source material, not general knowledge.
• Improvable: as sources are added or justifications refined, verdicts can change.

Version History