Main idea

While you work on a subject, future actions come up that have nothing to do with what you are doing. "We'll need to check that figure." "We should turn this post into a carousel." "Remember to proofread the English version." Left in the thread of the conversation, they produce one of two kinds of damage, never neither: either they are lost when the session closes, or they stay present and pull attention — yours and the assistant's — away from the problem at hand.

Parking them elsewhere settles both in a single gesture, and the second is the less obvious one. Writing "we'll need to check that figure" into a file of future actions, with an identifier and a status, isn't first of all about not forgetting it: it is about stopping holding it. As long as an action floats, it occupies part of what is doing the reasoning; deposited in a stable, named place, it occupies nothing and stays available.

It is the same gesture as everywhere else in the setup — pulling a piece of information out of a volatile medium to park it in a named place — but applied to what hasn't happened yet. And because this file is part of what gets reloaded at every opening, an action parked on Tuesday is still there on Friday without ever having been carried in between.

Why it matters

This requalifies an object usually taken for a tracking tool. A list of future actions isn't there to steer progress: it is there to unburden present reasoning of what doesn't concern it. The judgment you pass on it changes accordingly — a good list isn't the one that is up to date, it is the one you deposit into without hesitating.

It also gives a reason not to handle an action idea at the moment it arrives. Settling it straight away is precisely what you were trying to avoid: you interrupt the work under way for an object that wasn't ripe.

Nuances and limits

Depositing costs nothing, and that is its flaw. A list fed with no counterweight swells until it becomes illegible, and the file meant to free up reasoning becomes a weight loaded at every session — which is, incidentally, one of the signals that triggers a tightening of the workspace.

And not every action tolerates being deferred: a correction that invalidates what you are producing has to interrupt, precisely because parking it would mean carrying on from a false basis.

Open questions

• What distinguishes, at the moment it comes up, the action you can park from the one that has to interrupt?