The Neural Correlates of Consciousness

Don't ask "why is there feeling" — ask "the instant feeling appears, what changed in the brain"

In 1990, Francis Crick — co-discoverer of the DNA double helix — teamed up with the young neuroscientist Christof Koch and did something fairly heretical for the time: they argued consciousness should stop being the philosophers' private property and move into the lab. Their move was to sidestep the hard problem and chase a measurable target instead — the neural correlates of consciousness: the minimal set of neural activity jointly sufficient for one specific conscious experience ("right now I see this red dot"). Note both words, "minimal" and "specific": not the whole brain, just the handful of activity that's just enough to light this experience up.

Why does that phrasing make experiments possible? Because the vast majority of brain activity has nothing to do with consciousness. The brainstem activity regulating your heartbeat, digestion, and blood pressure never stops, and you feel none of it; even plenty of visual and auditory processing runs to completion in the dark (think of driving on autopilot — you still steer, still stop at lights, yet you weren't "watching"). So the real question becomes a piece of subtraction: take one stimulus, a trial where you were aware of it and a trial where you weren't; the difference in brain activity between them is the prime suspect for the consciousness part.

Take one image and let it flip back and forth in your eyes on its own

Subtraction sounds neat, but how do you guarantee "stimulus completely fixed, only consciousness changes"? In real life the moment the stimulus shifts, everything shifts with it, and you can't tell which slice of the difference is consciousness and which is the stimulus. Neuroscientists found a near-cheat loophole — binocular rivalry: show your left eye one image (say, a face) and your right eye a completely different one at the same time (say, a house). You'd expect a ghostly overlap of face and house, but the brain can't stand the contradiction, so it does something strange: it only lets you be aware of one at a time — first the face, then a few seconds later it snaps to the house, then back to the face, flipping endlessly, and you can't hold it still.

This becomes the clean lab bench neuroscientists dream of: what the eyes receive doesn't change for a second, yet what you consciously perceive keeps jumping between the two images. So the set of neural activity that flips along with your awareness gets cleanly divorced from the steady, unchanging stimulus. In the 1990s, Nikos Logothetis ran this on monkeys while recording single neurons: he found that the higher up the visual hierarchy you go, the more neurons flip with "what's seen" — in the earliest visual area, most cells still faithfully track the physical stimulus (which never changes, so they don't either); higher up, more and more cells switch to tracking the image the monkey is currently seeing. The trail of consciousness grows stronger the further up you climb.

The catch: the instant you "report" seeing it, you've poisoned the answer

For decades, experiments ran like this: we show you something, and when you see it you press a button (or say so) to report it. Perfectly natural, right? But around 2010, a group of people (Tsuchiya, Koch, and others) threw cold water on it: the act of "pressing to report" itself muddies the answer. Think about it — the path from "seeing" to "pressing" runs through a whole string of other jobs: gripping the experience in working memory, making a decision, sending a command to your finger. And all of those jobs lean heavily on the prefrontal cortex (prefrontal cortex).

Here's the problem: early experiments kept seeing "prefrontal lights up when you're conscious," so everyone reasonably concluded the prefrontal cortex is key to consciousness. But what if… the prefrontal glow isn't because you experienced something, but only because you were busy reporting it — a by-product of the "press" that got wrongly charged to consciousness's account? To test that suspicion, you'd need a way to drop the report yet still know which image you're seeing right now. That's the no-report paradigm.

How do you "know without asking"? Through the body's honest little tells. Show each eye stripes drifting in opposite directions, and whichever direction you're currently aware of, your eyeball involuntarily follows it in a fine twitch called optokinetic nystagmus — without you saying a word, the experimenter watches which way your eye jerks and knows which image you're "seeing." Pupil size and certain automatic reflexes can serve as the same kind of "sneak reader." The key result: with report removed, that patch of prefrontal "consciousness signal" shrinks substantially — a good chunk of it was the "report" being wrongly charged to consciousness's account.

This lit the "front vs back" debate that still hasn't died down: is the true home of consciousness up front in the prefrontal cortex (broadcasting, reporting, scheduling), or in the back, in the occipital-parietal-temporal sensory cortex that Koch and colleagues call the "posterior hot zone" (where the content of experience takes shape in the first place)? The no-report evidence nudged the scale toward "back," but it's nowhere near settled — a tension left perfectly poised for the heavyweight act coming next.

Make two sworn-enemy theories bet their own predictions in advance

Consciousness research has an old disease: theories galore, none able to convince the others — each can "explain" the existing data eloquently, because the explanations are patched in after the fact. Philosophy of science said long ago that real skill isn't explaining prettily, it's daring to commit in advance to the condition under which you'd lose. So starting in 2019, something hardcore happened: proponents of two major consciousness theories sat down at one table and signed an adversarial collaboration — the Cogitate consortium.

The two colliding sides: Global Neuronal Workspace Theory (GNWT, the Dehaene camp — next issue's star), which says a piece of information only reaches your awareness once it's "ignited" and broadcast across the brain (especially the prefrontal cortex); and Integrated Information Theory (IIT, the Tononi camp), which says consciousness is information being highly integrated and sustained in the posterior cortex, with little to do with whether the prefrontal cortex broadcasts. Crucially, both sides wrote down their predictions in advance, in black and white, along with an agreed "this result counts as my loss," and handed it to neutral labs to run the same subjects and the same stimuli — nobody gets to change their story afterward.

The result — announced in 2023, formally published in Nature in 2025 — was a "thrilling match with no winner." On one hand, activity in the posterior cortex really did let researchers decode quite well what you were seeing — a plus for IIT's emphasis on "the back"; yet IIT's signature hard prediction, that posterior regions would stay synchronized and hold the experience up continuously, was not supported by the data. On the other hand, some information did reach the prefrontal cortex, a consolation for GNWT; but GNWT's heaviest bet — that at the instant the stimulus vanishes, the prefrontal cortex would ignite again to broadcast — came up empty too. Both sides tripped over their own predictions, and neither could land the knockout.

That result sounds like "all for nothing," but it's exactly the opposite. For the first time it let a consciousness theory genuinely lose — not by verbally backing down, but by having a piece of its own signed-and-sealed prediction falsified on the spot. This is what it looks like to drag consciousness out of speculation and into science: not whoever shouts loudest wins, but whoever dares to bet and, on losing, admits it. Next issue we pull one side of this collision out on its own — the global workspace — and see what gives it the nerve to say "consciousness = ignition + brain-wide broadcast."