Theories · about 5 minutes
The Stage Door
Your eyes are open. The lights are on. Something is displayed in front of you, large and clear, for a tenth of a second -- the same tenth of a second in which you are perfectly capable of reading anything else.
You will not see it. Not "you'll see it and forget" -- you will look straight at it and there will be nothing there. Let's do it to you eight times so you can't blame the first one on bad luck.
Before we start. This exhibit flashes letters in one spot, quickly -- about nine a second. The background never changes brightness and nothing strobes, but if fast visual change is a problem for you, there is a slower setting below. Use it: at the slower speed the effect disappears completely, and that turns out to be the whole point.
Trial 0 of 8
+
Two digits went past. Which ones, in order?
Your results
What just happened
This is the attentional blink, first named by Raymond, Shapiro and Arnell in 1992, and it is one of the tidiest facts in cognitive science. Find a first target, and for roughly the next 200 to 500 milliseconds a second target is very likely to vanish -- not blur, not half-register. Vanish. Wait past about half a second and it comes back as if nothing had happened.
It is called a blink because it behaves like one, but nothing physical closed. Your eyes were open the whole time and never moved. Whatever shut, shut further in.
And here is the part that turns a curiosity into evidence. That missed digit was not ignored by your brain. When this is run with words instead of digits and the brain is wired up, a word you completely fail to report still produces the electrical signature of a word whose meaning has been understood. The processing happened. The recognition happened. The reaching-you did not.
The theory this room is named for
Bernard Baars proposed in 1988 that consciousness works like a theatre. Enormous amounts of your brain are running in parallel, all the time, and almost none of it is conscious -- the sound of the fridge, the position of your left foot, the grammar of this sentence. These are the actors backstage. They are all working. They are all competing. And there is one spotlight.
What wins the spotlight gets broadcast -- made available to everything else at once, so you can report it, remember it, act on it, be reminded of your grandmother by it. That global availability, on this theory, is consciousness. Not a place in the brain. A publication event. Stanislas Dehaene's version, the global neuronal workspace, adds the physiology: an all-or-nothing "ignition" of widespread frontal and parietal activity about a third of a second after the stimulus, which either happens or doesn't.
Everything is running. One thing at a time gets out. You are the audience, and also the theatre, and also -- this is the awkward bit -- the play.
The blink, on this account, is the door being busy. Your first digit is mid-broadcast, occupying the machinery that makes things reportable, and the second digit arrives at a door that hasn't finished with the last customer. It gets seen, processed, understood, and never published. By the time the door is free, the moment is gone.
Now go and run the slow version, if you haven't -- it's the control, and it is worth two minutes. Space the same letters out and the blink does not shrink, it disappears; every digit sails through. The effect isn't a limit on how much you can see. It's a limit on how fast you can consciously see, and it has a duration -- which is a very strange property for a thing to have.
Why it matters, and where it stops
Global workspace is currently one of the two or three leading scientific theories of consciousness, and it has the great virtue of predicting things that then happen. It explains why you can't listen properly to two conversations, why a thought can be on the tip of your tongue, and why the blink lasts about as long as it does.
In 2023 it was put in a ring with its main rival. A large adversarial collaboration -- theorists on both sides agreeing in advance what results would count against their own position, which is a rarer act of scientific courage than it sounds -- tested global workspace against integrated information theory. Both took damage. Neither was killed. That is roughly the healthiest outcome available, and it is why the needle below sits where it does.
But stay with the theory a moment longer, because it walks straight into the wall this whole building keeps hitting. Suppose it's completely right. Suppose consciousness is exactly what happens when information wins the competition and gets broadcast widely. Fine: why is being broadcast something it feels like anything to undergo? A memo sent to every department of a company is not thereby enjoyed by the company. Global workspace explains, beautifully, which information you can report on. It does not explain why there is anyone home to report it to.
That question has its own room, and it's back down the corridor: Exhibit № 24.
Take it further
Bernard Baars laid it out in A Cognitive Theory of Consciousness (1988); Stanislas Dehaene's Consciousness and the Brain is the readable modern version with the brain imaging attached, and is very good on how you test a thing like this at all. The phenomenon itself has its own literature under attentional blink, and the theory under global workspace theory.
One honest note about what you just did: eight trials on one person in a museum is a demonstration, not an experiment. The real effect is a population average over many trials, and any individual can get lucky, blink less than average, or simply be having a good day. If it didn't happen to you, that isn't a failure -- of you or of the effect. It's what a sample size of eight looks like.