Seven scientific theories of consciousness, four wider positions, and how each is holding up now that the theories are being tested against each other in public.
Last reviewed November 3, 2026 · Updated when results land
How to read this. Odds are our editors’ confidence that a theory is closest to right, not a market. Changelog. Nov 2026: Dendritic Integration (4%) and Orch-OR (1%) added to the board, others scaled to make room; preliminary NYU results nudge RPT down. Apr 2025: COGITATE 1 published, IIT 28 to 22, GNWT 26 to 20, predictive processing 16 to 18.
The theories disagree about more than brain regions. Their answers decide whether your chatbot could feel anything, which animals count, and what doctors owe a patient who can’t respond.
Simplified. Each theory’s answer is how its main proponents would put it today; individual researchers disagree.
The theories are mostly built from lab perception tasks and patients. The states that people who meditate or take psychedelics report are starting to become evidence too.
Carhart-Harris and Friston’s REBUS model says psychedelics loosen the brain’s highest-level predictions, so bottom-up signals get more say. It’s the clearest case of a consciousness theory making claims about altered states, and much psychedelic research is now framed in its terms.
Carhart-Harris & Friston, Pharmacological Reviews 2019 ↗Ruben Laukkonen, Karl Friston and Shamil Chandaria propose that consciousness arises when a predictive model includes a model of its own knowing, and use meditative states, including the fading of the sense of self, as test cases.
Laukkonen, Friston & Chandaria, Neurosci Biobehav Rev 2025 ↗Advanced meditators describe “cessations,” moments where experience seems to switch off and come back. A Harvard-affiliated team has recorded them with EEG in an intensively studied expert. Any theory has to say what changed in those seconds.
Chowdhury et al., Neuropsychologia 2023 ↗In 2025, Cleeremans, Mudrik and Seth reviewed the state of the field and called for more phenomenology, careful descriptions of what experience is like, alongside tougher, theory-driven tests.
Cleeremans, Mudrik & Seth, Frontiers in Science 2025 ↗124 researchers signed an open letter calling integrated information theory pseudoscience, arguing its core claims are untestable and its implications too far-reaching. In April 2025 Nature Neuroscience ran all three sides in one issue: the critics, the IIT group’s reply, and a comment from Gómez-Marin and Seth calling the charge unwarranted.
In 1998 Christof Koch bet David Chalmers a case of wine that the neural correlates of consciousness would be found within 25 years. At the 2023 ASSC meeting, Koch conceded and handed over the wine. He immediately proposed a new bet for 2048.
Two back-to-back papers argued over where consciousness lives. Boly, Koch, Tononi and others made the case for the posterior cortex. Odegaard, Knight, and Lau argued the prefrontal cortex is required and the lesion evidence has been misread.
The first large adversarial collaboration: seven labs, 256 participants, three recording methods, predictions registered in advance by IIT and GNWT. The consortium’s summary: the results fit some predictions of each theory and substantially challenged key tenets of both. Hakwan Lau, who reviewed it for Nature, argued the design didn’t put either theory to a real test.
Which theory is closest to right? Pick one, see how your call compares with our editors’ odds, and cast your vote. We’ll publish what readers think.
Pick a theory to see our odds.
The consciousness science room: the theories, the adversarial tests, hidden awareness in patients. Peer-reviewed, last ten years.
Enter the room → Ask the researchOur AI answers only from peer-reviewed papers in the library and cites every claim.
Ask it → LibraryStart with the people behind these theories, or ask the Guide where you should begin.
Open the library →Researchers: if we’ve mischaracterized your view, tell us and we’ll correct it.