The Searchlight: Attention and Environment

You sat down to work at nine. It's now half eleven. In that time you didn't leave the desk, didn't sink into a series, you were apparently here the whole while.

Done: a page and a half.

Where two and a half hours went isn't obvious, because they weren't stolen in large pieces. They were carried off thirty seconds at a time: glance at a notification, reply in a chat, check email, look up one thing and come back. Every time you came back honestly. You just came back to somewhere other than where you left.

A searchlight in a dark hangar

Picture an enormous dark hangar and a single searchlight on a movable mount. Everything lit is visible to you and you can work with it. Everything else exists but doesn't concern you.

That's how attention is built: not general lighting but a narrow spot you can move.

This searchlight has three properties, and all three are unwelcome.

There's one spot. You can't light two places at once. You can swing the beam back and forth quickly, creating the illusion that everything is lit, but at any given moment it shines on one point.

The mount is heavy. The beam doesn't jump instantly: aiming it somewhere new means turning the mechanism and then adjusting the focus. Every turn costs time, and it costs it twice — on the way out and on the way back.

The lamp heats up. You can't burn a searchlight indefinitely: after an hour or so the brightness drops, and that isn't laziness, it's a resource.

Attention isn't lighting, it's a searchlight. One spot, and turning it costs money.

Multitasking doesn't exist

More precisely, it exists, but not the kind people take it for.

The brain doesn't perform two tasks requiring thought in parallel — it switches between them, and every switch is charged for. Robert Rogers and Stephen Monsell measured this in pure form back in 1995: people worked either on one type of task or alternated between two. Even when the order was known in advance and they could prepare to switch, alternating was consistently slower. The cost of the transition never went away.

In real work it's far higher than in the lab, because the residue from the previous task is added on top. Sophie Leroy described an effect she called "attention residue": when a person abandons a task unfinished and moves to another, part of their attention stays with the previous one — and the quality of work on the new one drops. That's exactly why "just a second to glance at the chat" doesn't cost a second: the chat leaves the screen but not the head.

A separate and especially expensive variety is interruption. Gloria Mark and colleagues observed office workers and found that after an interruption a person is far from quick to return to the original task — other business wedges itself in between leaving and coming back. And in another study the same group showed something curious: when people were pushed to work faster because of interruptions, they did indeed keep up, but paid for it in increased stress and strain.

The practical conclusion: fragmented attention isn't a slower version of focused attention but a qualitatively different, worse mode of work. An hour and a half with notifications isn't equal to an hour and a half without them: it isn't "slightly worse," it's worse in kind.

The phone steals even when it's off

The most counterintuitive part of the subject.

Adrian Ward and colleagues (2017) tested a simple thing: does the mere presence of a smartphone affect cognitive ability? Participants performed working memory and attention tasks under three conditions: phone face down on the desk, phone in a pocket or bag, phone in another room. All phones were silenced — no notifications.

Those whose phone was in another room did best. And the participants themselves didn't think the phone had hindered them in any way.

The authors' explanation: to not look at the phone you have to actively not look at the phone, and that's work which occupies part of that same limited resource. The searchlight has to be constantly held back from turning — and holding it takes power.

Hence the rule that saves the most effort for the least effort: don't fight the distraction, remove it physically. A phone in another room requires no willpower. A phone on the desk requires it continuously.

The environment decides for you

The same is true of everything else around you. The environment isn't background but a set of decisions made in advance.

Remove from sight whatever the beam will drift towards. Not only the phone: extra tabs, a second loaded task on the desk, a messenger open on the second monitor. Everything that can be closed should be closed before you start.

Separate the place of work from the place of rest. Physically if you can, or at least by a different chair or a different corner of the table. The brain quickly binds behaviour to setting; a place where you both work and scroll sends no signal at all.

Prepare everything beforehand. Materials, water, the task list. Every trip to fetch something is a legitimate excuse for the beam to swing away.

Come to an arrangement with people. A considerable share of interruptions in a person's life aren't notifications but other people. A known time when you can't be disturbed works better than the hope that nobody will disturb you.

Noise. Silence or a monotonous background. Speech in a language you understand is the worst possible background: the brain processes it automatically and occupies that same auditory part of working memory the article on the workbench was about. Music with lyrics interferes with reading more than instrumental music for the same reason.

Rhythm: working in blocks

The lamp heats up — which means you have to work in cycles rather than until victory.

The best-known scheme is 25 minutes of work and a 5-minute break; it came up in the article on procrastination. But there it solved the problem of getting started, and here it solves a different one: staying in. A block with a known end is easier to work through without switching, because every temptation gets the answer "after the timer."

A few practical refinements:

The length adapts. 25 minutes is a good starting point, but once drawn in, many people work better in blocks of 45–90 minutes. The guideline is simple: finish before you start falling apart, not after.
The break has to be a real one. Social media during a break doesn't restore the searchlight, it keeps driving it. The opposite works: stand up, walk, look into the distance, be quiet. There'll be an article on the aquarium about this — ten minutes of rest after material yields a measurable gain.
Write down whatever intrudes. Keep a sheet beside you for stray thoughts: "buy tickets," "reply to Vanya." A thought written down stops occupying space on the workbench, and has no reason to keep intruding.
When interrupted, mark the point. If you do get pulled away, spend ten seconds writing down exactly where you stopped and what you were about to do. That sharply reduces the cost of coming back.

Attention is consumable but restorable

The last thing worth understanding: the ability to concentrate is a state, not a character trait.

It runs down over the day, and runs down faster with the number of decisions made and impulses suppressed. It's restored — by sleep, movement, silence, and, judging by Marc Berman's research that the article on the body will cover, better by a walk among greenery than a walk along a road.

Two practical conclusions follow. First: put the hardest thing at the start of the day, when the lamp is bright, not in the evening, when it only stretches to email. Second: if concentration isn't coming today, the question isn't "what's wrong with me" but "when did I last sleep, eat, move and sit in silence."

In practice

Phone in another room, not in your pocket. The cheapest action with the largest effect in this whole article.

One task in view. Everything superfluous closed before work starts, not "minimised as needed."

Work in timed blocks. Start with 25 minutes, tune it to yourself. Inside a block, not a single switch, whatever happens.

Keep a sheet for intruding thoughts. Written down — back to work.

Schedule the hard things for your peak and routine for the dip. Decide in advance which block of the day goes to what — a decision easier to make with a fresh head than a tired one.

Make the break a break. Stand up, step away, look into the distance, don't grab a screen.

Check yourself

Close the article and answer in your own words:

What are the three properties of the attention "searchlight," and what follows from each?
Why does "just a second to glance at the chat" cost far more than a second?
What did Ward's study with the phone in another room show, and why does it work even without notifications?
Why is speech in a language you understand the worst kind of background noise?
What should you do in the first ten seconds after being interrupted?

In short

Attention is a narrow moving spot, not general lighting: one place at a time, a turn you pay for, a lamp with a limited resource.
There's no multitasking in thinking, there's switching — and it costs time plus the attention residue from the abandoned task.
The mere presence of a phone in view lowers results even when it's silent: holding yourself back from looking is work too.
Don't fight distractions with willpower, remove them in advance. The environment is a set of decisions made with a fresh head.
Work in blocks with real breaks; write out stray thoughts; when interrupted, mark your point of return.
Concentration is a state, not a character trait: it runs down over the day and is restored by sleep, movement and silence.
Attention isn't lighting, it's a searchlight. One spot, and turning it costs money.