The Journal

Write it down badly: the case for a messy brain dump

Many household jobs come to mind at a bad moment: hands full, halfway out of the door, mid-sentence. Writing one down properly at that moment means deciding where it goes, when it is due and whose it is, so it often does not get written down at all. The fix is to split the job in two: capture badly now, sort it later at a time you have booked.

Published 2 November 2026

---

Two jobs that look like one

You notice the smoke alarm chirping while carrying a basket of washing up the stairs. Writing that down properly means choosing a list, a date, a time, maybe a person, and perhaps a note about which alarm it was. That is five decisions at the moment both hands are full. So the job stays in your head, where it chirps for three more days.

Two different jobs are hiding inside "write it down". Capture gets the thing out of your head and into one trusted place. Organising decides what it is, when it happens and who does it. Capture has to be fast, because it happens at the moment of noticing, and that moment is rarely convenient. Organising has to be careful, because it is where a vague worry becomes something you can start. A system that asks for both at once taxes the fast job with the cost of the slow one.

This piece argues for doing them separately: capture that is deliberately messy, and a short tidy booked for later. The research below explains why capture is worth doing at all. It does not test this exact split, and the piece says where it stops.

What the offloading research actually shows

Psychologists call writing something down so you no longer have to hold it in mind cognitive offloading. Evan Risko and Sam Gilbert, in a 2016 review in Trends in Cognitive Sciences, define it as "the use of physical action to alter the information processing requirements of a task so as to reduce cognitive demand", and one of their examples is setting a phone to remind you of an appointment.

Three findings matter here.

Offloading frees memory for the next thing. Storm and Stone, in Psychological Science in 2015, reported three experiments in which saving one file before studying a second "significantly improved memory for the contents of the new file". The effect did not appear when saving was made to seem unreliable, or when the first file was too slight to interfere with the second. Both conditions matter at home: the place you capture into has to be one you trust, and the gain is largest when you are carrying a lot.

External reminders are far more accurate than memory, in the laboratory. A 2023 review by Gilbert and colleagues in Psychonomic Bulletin and Review describes their standard task, in which people hold several delayed intentions while doing something demanding. Using their own memory, people were typically around 50 to 60 per cent accurate; with external reminders, typically 90 to 100 per cent. The same review notes that a reminder has a cost, "the time and effort setting it up", and that people in these studies tended to set more reminders than was strictly optimal.

An unfinished job keeps intruding until there is a plan for it. E.J. Masicampo and Roy Baumeister, in a series of studies published in the Journal of Personality and Social Psychology in 2011, found that an unfulfilled goal caused intrusive thoughts during an unrelated reading task and worse performance on an unrelated puzzle. When participants made a specific plan for the goal, those effects went away; what removed them was a plan. These measured intrusion and interference, not memory for the unfinished task, which matters because the older claim that unfinished tasks are remembered better did not hold up in a 2025 meta-analysis, as The small job that costs more to remember than to do sets out. This piece found no published replication of the 2011 studies when it looked on 2 November 2026.

The limit is plain. None of these studies looked at household lists, shared lists or voice capture. The 2023 review itself says "we do not yet know much" about how offloading measured in the laboratory relates to how people offload in everyday life. What follows is reasoning from these results, not a tested method.

The cost of a decision at the door

If a reminder only helps once it exists, then the effort of creating it decides how many reminders exist. Every field a capture tool asks you to fill in is a reason not to capture. A list app that opens on a form with a date picker, a project menu and a priority flag has made capture as expensive as organising, and at the doorway it loses to "I'll remember".

So the test of a capture method is how little it asks. A scrap of paper on the fridge passes. A note to yourself in a messaging app passes. Saying it out loud to a phone passes, if what you said turns into a task without further questions.

This is where Barriers tries to sit. A task can be spoken rather than typed on iPhone, iPad and Mac. Saying several jobs in one breath and having each land as its own task is part of the paid plan, offered free for a limited time the first time you use it. On the same devices it can read a photo of a note into a task. Neither asks you to decide anything at the moment of capture beyond what you said. The date, the person and the list can wait for the tidy.

Bad capture beats a good intention

A messy note says "smoke alarm chirp". A tidy one says "Replace the battery in the landing smoke alarm, Saturday morning, buy 9V batteries first". The tidy one is better to act on. The messy one is better to have, because it exists.

The point of capture is not to describe the job well. It is to stop the job living in your head, where the 2011 studies suggest it keeps costing attention until something is settled about it. A three-word note does not settle it, but it holds the job somewhere safe until the tidy does.

So lower the standard on purpose. No capitals, no dates, no categories. A half-sentence that will make sense to you on Sunday is enough. If you find yourself pausing to word it properly, you are organising, and the moment for that is later.

The tidy is a separate appointment

Capture without a tidy is a pile, and a pile becomes a second thing to avoid. The tidy is where each scrap becomes a line you could start: one physical action, a day, and a name if it is not yours. Why starting is harder than doing sets out how to write that line. It is also where the plan in Masicampo and Baumeister's sense gets made, and it was the plan, not the note, that quietened the intrusions in their studies.

Book it: ten minutes at a fixed point in the week, the same day each time. How long a household needs is a guess, not a finding; start with ten minutes and adjust. Is decision fatigue real? makes the related case for settling a recurring household decision once, at a quiet time, instead of every week under pressure. Some scraps will turn out to need nothing: a job that has fixed itself, or one nobody is going to do. Deleting those is a legitimate result of the tidy, not a failure of it.

One inbox, or the note is lost anyway

Storm and Stone's effect disappeared when saving was unreliable, and the household equivalent is capture spread across six places: two notebooks, the back of an envelope, a voice memo, a message to a partner and the fridge door. Each one is fast. Together they are untrustworthy, and the tidy becomes a search.

Pick one place to capture into, or at most one place per person, and make the tidy empty it. If capture has to happen on paper sometimes, the tidy is when paper goes into the main list. If two adults share the household, the capture place does not have to be shared, but what comes out of the tidy should be, so that a job with someone's name on it is visible to them. Naming a person is only the start of handing a job over; whether they have seen it and agreed is a separate step.

What not to capture at all

Three kinds of thing do not belong in the capture place.

Things you will do in the next two minutes. Capturing them costs more than doing them. The small job that costs more to remember than to do looks at where that rule comes from and how far the evidence for it goes.

Things that are already scheduled elsewhere. A dentist appointment that lives in the calendar does not need a second copy as a task, and two copies drift apart.

Worries that are not jobs. "Money feels tight" cannot be done. If there is a job inside it, such as "look at last month's statement", capture that instead. If there is not, a list is the wrong place for it, and putting it there teaches you to stop trusting the list.

---