Attention Residue

You did everything right and the hour was still garbage.

Door shut. Phone in a drawer, face down, in another room if you're the type. Notifications off, tabs closed, a clean desk and ninety protected minutes you fought someone to get. This was the setup you have been told for a decade is the answer, and you built it, and then you sat down inside it and produced forty minutes of nothing followed by a paragraph you deleted.

Nothing interrupted you. That is the part that should have told you something.

So you reached for the only explanation left, which is that the problem must be you — your focus, your discipline, some deficiency of attention that persists even after you removed everything that could compete for it.

Nothing competed for it because the seat was already taken. You walked into that room carrying the previous thing, and it does not put itself down because you closed a door.

CHAPTER I: THE THING YOU BROUGHT WITH YOU

In 2009, Sophie Leroy published a paper in Organizational Behavior and Human Decision Processes with a title that reads like something muttered at a desk: "Why is it so hard to do my work?"

Her subject was the moment of transition. Not interruption itself — the switch. What actually happens in the seconds after you stop Task A and start Task B.

The assumption everyone runs on is that attention is a beam. It points at one thing; you move it; now it points at another. Under that model the cost of switching is the moving, a second or two of reorientation, and then you are fully present on the new thing.

Leroy's finding is that attention does not move cleanly. A portion of it stays behind, still running on the previous task, still processing it — and it does this while you are consciously working on something else and would report, if asked, that you were focused. She called it attention residue.

Across two experiments she showed the consequence: when people transition away from an unfinished task, they find it difficult to stop thinking about it, and their performance on the next task suffers.

Read that carefully, because it relocates the damage. Most of what you have read about interruption is about the interrupted task — the thing you had to leave, the context you have to rebuild when you come back. Leroy is measuring the opposite direction. The cost is being paid by the task you switched to. By the important thing. By the ninety protected minutes.

Which is why removing distractions did not save your morning. The distraction had already happened, at nine forty, and you brought it into the room with you at ten.

CHAPTER II: FINISHING IS NOT ENOUGH

The obvious remedy is completion. If unfinished tasks are what stick, finish things before you move on. Close the loop, get to a stopping point, and arrive clean.

Leroy tested that, and this is where the paper stops being intuitive.

Her own summary, in the abstract, is blunt: being able to finish one task before switching to another is not enough to enable effective task transitions.

You can complete a task and still carry it. Getting to the end does not, by itself, release the cognitive hold. Anyone who has finished a difficult conversation and then spent an hour re-running it already has the data.

So what does release it?

Leroy's answer: time pressure while finishing. Her wording is that time pressure while finishing a prior task is needed to disengage from it, and that it contributes to higher performance on the next task.

Being rushed on Task A makes you better at Task B.

That sounds wrong and it is worth understanding why it isn't. The mechanism is not that hurry improves the work. It is that hurry forces a decision. When time is abundant, finishing is soft — you close the document while still turning it over, still considering the alternative phrasing, still holding it open in case. There is no moment of this is now concluded, so nothing signals the mind to release the resources. When time is short, you are compelled to reach a state of closure and commit. The task gets shut rather than trailed off.

Closure is not the same as completion. Completion is a fact about the work. Closure is an event in you, and it is possible to have the first without the second — which is precisely the state most of your days are made of.

And now notice what this does to the advice you have been following. The standard prescription is to protect generous, unhurried blocks. Leroy's result suggests the generous unhurried block, ended vaguely, is a residue factory: you drift out of it without ever concluding, and you carry it into whatever comes next. The deadline you resent may be doing more for your afternoon than the calm you're protecting.

CHAPTER III: THE ONE YOU HAVEN'T GOT BACK TO YET

Nine years later, Leroy and Theresa Glomb published a second paper, in Organization Science, that identifies what keeps the hold locked.

They looked at people interrupted mid-task and manipulated one thing: what those people expected about returning. Specifically, whether they anticipated being under time pressure when they resumed the interrupted work.

Participants who thought they would be rushed on their return showed high attention residue and performed significantly worse on the interrupting task. Not vaguely worse. They did not process information carefully. They failed to notice errors. When asked to make a decision from what they had taken in, they were less likely to find the optimal solution.

None of that is about the task they'd left. It is a description of a man doing his current work badly while a different, unresolved thing runs in the background.

Put the two papers side by side and the mechanism is one thing seen from two angles. Rushed finishing produces closure and frees you. Anticipated future rush denies closure and holds you. In both cases the variable is not how much work is left. It is whether the matter has been settled in your head.

Which explains the specific texture of a bad day better than any theory of distraction. The reason you cannot concentrate at two in the afternoon is rarely that something at two is competing. It is the thing at eleven that you left in an unresolved state, with a vague sense that it will be difficult when you get back to it, and that vague sense is doing continuous background work at some cost to everything you attempt in between.

It also explains why the worst residue comes from the things you are dreading. A task you expect to be unpleasant on resumption is a task your mind will not set down, so it taxes every hour between now and then. You pay for the difficult conversation twice: once when you have it, and once across all the hours you spend not having it yet.

CHAPTER IV: A DAY MADE OF ARRIVALS

Now put this next to how a working day is actually shaped.

You do not have one transition. You have thirty. Meeting to desk, desk to message, message to call, call to the thing you were supposed to spend the day on. Each one is a switch, and almost none of them end in closure — meetings run to the minute and stop mid-thought, messages get half-answered, calls end with something unresolved that you will "pick up later."

So the residue does not clear between events. It layers. By mid-afternoon you are not carrying the last thing; you are carrying six things, each unsettled, each drawing something, and the important work is being attempted by whatever fraction of you is left over.

Then measure yourself against that and draw the obvious wrong conclusion. This is where the diagnosis goes bad, because the felt experience — sitting at a clean desk, unable to think — is genuinely indistinguishable from being bad at concentrating. There is no sensation that tells you your capacity is committed elsewhere. You just find the work hard and yourself insufficient.

And the compounding is that the conclusion changes what you try. Believing the problem is focus, you buy more isolation: longer blocks, stricter rules, another app. Each attempt fails for the same reason the first one did, because none of them touch the loose ends upstream, and each failure is further evidence about your character. Men have spent years on this loop and ended up with a settled belief that they cannot concentrate any more, which was never once the thing being measured.

CONCLUSION: CLOSE THINGS, DON'T JUST FINISH THEM

Take the range seriously first. This is two laboratory experiments in the 2009 paper and controlled studies in the 2018 one, in a field where effects are often smaller and less general than the headline. And the time-pressure result should be handled with care: it is about a bounded rush that forces closure at the end of a task, not an argument for living under pressure. Sustained high demand with no control over how you meet it is the arrangement that shows up in the mortality data, not a productivity technique.

Within those limits, four things follow, and the first one is the one with the evidence behind it.

Write a ready-to-resume note before you switch. This is Leroy and Glomb's own intervention and it takes about a minute: before you leave a task, write down exactly where you are and what you had planned to do next. Not a to-do item — the actual state and the actual next move. The point is not the note's future usefulness. The point is that writing it lets you cognitively set the task down, which is the thing you have never been doing.

Give yourself less time, not more. Put a hard stop on the work rather than an open runway, and let the last stretch be genuinely rushed. The finish will be less elegant than you'd like and the following hour will be markedly better. If you only take one behavioural change from this, take the hard stop.

Never put the important thing directly after the unresolved thing. Sequence matters more than duration. Ninety minutes at eleven, immediately after a meeting that ended ambiguously, is not ninety minutes. Either close the meeting properly — five minutes, decisions stated, next step named, out loud — or put something disposable between it and the work that matters.

And read your own signal correctly. When you notice you are rehearsing a conversation, re-running a decision, or drafting a reply in your head, that is not distraction and treating it as a discipline failure will not touch it. It is an open loop announcing itself. The move is not to push harder into the current task; it is to spend sixty seconds closing the old one — decide it, note it, or schedule it — and then return.

You built the perfect room and had a terrible morning in it, and concluded the fault was yours.

You were never alone in there.

SOURCES

Primary: Sophie Leroy, "Why is it so hard to do my work? The challenge of attention residue when switching between work tasks," Organizational Behavior and Human Decision Processes 109(2), July 2009, pp. 168–181 — the concept of attention residue; two experiments showing that people find it difficult to transition attention away from an unfinished task and that performance on the subsequent task suffers; and the abstract's own statements that being able to finish one task before switching is not enough to enable effective task transitions, and that time pressure while finishing a prior task is needed to disengage from it and contributes to higher performance on the next task.

Primary: Sophie Leroy and Theresa M. Glomb, "Tasks Interrupted: How Anticipating Time Pressure on Resumption of an Interrupted Task Causes Attention Residue and Low Performance on Interrupting Tasks and How a 'Ready-to-Resume' Plan Mitigates the Effects," Organization Science 29(3), 2018, pp. 380–397 — participants anticipating time pressure on resumption showing high attention residue and significant performance decrements on the interrupting task, including failure to process information carefully, failure to notice errors, and reduced likelihood of identifying the optimal solution; and the ready-to-resume plan, in which briefly noting where one was and what one planned to do next before switching mitigates the effect.

Related within this series: red-cardinal-trained-interruptibility.md (Dabbish, Mark & González, 2011) covers why the switches happen; this piece covers what each one costs on arrival. red-cardinal-the-status-gradient.md is the reason the time-pressure finding is fenced in here rather than generalised into advice about working under pressure.

Both papers are experimental work in organisational psychology with the usual limits on sample and generalisability. The time-pressure result concerns a bounded rush producing closure at the end of a task and is not evidence that sustained time pressure improves performance; the text states this explicitly.

RELATED

• Trained Interruptibility
• Just Think — forthcoming, week 4
• Enterprise AI Finally Arrived