Attentional Blink

The brain can see a letter and then go briefly blind because it is busy reporting the letter it just saw. In Raymond, Shapiro, and Arnell's 1992 RSVP task, people often missed a second target when it appeared about 200-500 milliseconds after the first. The eyes worked. The display worked. The bottleneck was attention turning perception into report.

How it works

The standard experiment is blunt: show 10-20 items per second in the same place, ask for two targets, then vary the gap between them. If T2 appears right after T1, people often catch both. If T2 arrives a few hundred milliseconds later, accuracy drops. By roughly 700 milliseconds, performance recovers.

That odd pattern is the blink. It is not a slow eye movement problem because the stream sits in one spot. It is not pure memory failure because the deficit depends on timing. The working idea is that T1 enters a limited consolidation stage, and T2 arrives while that stage is occupied.

Chun and Potter's 1995 two-stage model captures the basic shape:

The sharp line is this: consciousness is not the same as sensory registration. A stimulus can enter the system and still fail to become something a person can say they saw.

The timing

Lag after T1	Typical result	What it suggests
0-100 ms	Often spared	T1 and T2 may be grouped together
200-500 ms	T2 often missed	report-making is occupied
600-700+ ms	Recovery	consolidation window has reopened

The phrase "attentional blink" is slightly misleading. Nothing closes like an eyelid. The system keeps sampling, but the route from sampled signal to reportable object narrows.

This matters outside lab displays. Air-traffic screens, radiology scans, code review, trading dashboards, and phone notifications all punish the same assumption: that seeing fast events is the same as registering them.

What's contested

The effect is settled; the explanation is not. Some models treat the blink as a hard bottleneck in working memory consolidation. Others argue it comes from control settings: the brain over-invests in protecting T1, then suppresses nearby distractors and accidentally suppresses T2.

There is also a training question. People can reduce the blink under some conditions, and task-irrelevant activity sometimes weakens it, as Olivers and Nieuwenhuis reported in 2005. That makes the blink less like a fixed machine limit and more like a policy the brain adopts under pressure.

Why this touches other realms

Attentional blink belongs next to concept information theory because both care about bottlenecks, not vibes. Shannon's 1948 paper stripped communication down to signal, noise, and channel capacity. The blink asks a biological version of the same question: how much event-rate can a mind carry before the second signal stops becoming an object?

It also links to concept fermi paradox in a stranger way. The Fermi question asks why a universe full of possible signals looks silent. Attentional blink asks how a visual field full of real signals can still produce silence in the report. Both warn against equating absence of report with absence of event.

An open question

If attention blinks because it protects the first target, what is the cost of a mind trained never to protect anything for long? The next page here is concept working memory, because the blink is where perception starts paying rent to memory.

Key sources

Raymond, Shapiro, and Arnell, 1992, "Temporary suppression of visual processing in an RSVP task: An attentional blink?", Journal of Experimental Psychology: Human Perception and Performance - the naming experiment.
Broadbent and Broadbent, 1987, "From detection to identification: Response to multiple targets in rapid serial visual presentation", Perception & Psychophysics - early target-detection groundwork.
Chun and Potter, 1995, "A two-stage model for multiple target detection in rapid serial visual presentation", Journal of Experimental Psychology: Human Perception and Performance - the canonical two-stage account.
Dux and Marois, 2009, "The attentional blink: A review of data and theory", Attention, Perception, & Psychophysics - theory comparison after 17 years of work.
Martens and Wyble, 2010, "The attentional blink: Past, present, and future of a blind spot in perceptual awareness", Neuroscience & Biobehavioral Reviews - later map of the field.

Abhishek's take

What grabs me here is that the failure lasts less than half a second, yet it breaks the naive story that attention is just effort. The brain is not a camera with willpower. It is a queue, and queues lose packets when arrival rate beats service time.

Tags: #attention #cognition #perception #neuroscience #consciousness

Abhishek