Starting · cornerstone
What a deep-work ritual actually looks like
Deep work is usually pictured as the timer and the headphones. The part that decides the hour happens before the timer starts, and the part that decides tomorrow happens after it stops.
When people think about deep work, they usually picture this part.
Sit down. Start a timer. Put on some music. Work.
That picture is not wrong. It is just the middle of something. A deep-work ritual actually starts before the timer, and it does not end when the timer rings. There are four parts, and most tools, and most of us, only plan for one of them.
The four parts
The framework Ember is built on is short enough to remember on a bad day: Ignite, Focus, Recover, Return. Enter, hold, leave, and set up the next entry.
Ignite: decide what this block is for
Before anything opens, answer two questions in writing. What will be different when this block ends? And what am I deliberately not doing until then?
This is the step that the timer-and-headphones picture skips, and it is the one with the best evidence behind it. Gollwitzer’s work on implementation intentions found that people who form a specific plan about when and where they will act (“if it is 9 a.m. and I am at the desk, then I open the draft”) act on it far more often than people who merely intend to.1 The effect is on starting, not on brilliance. That is fine. Starting is the part that fails.
The “not doing” half matters as much as the goal. Naming one thing you are refusing for the next block, whether the inbox, the phone or the second tab, turns a vague hope of not being distracted into a decision that was already made.
Focus: one task, one protected block
One meaningful task. One protected block of attention. Fewer decisions about where your mind should go next.
The reason to be strict about one is not discipline for its own sake. Leroy’s studies on attention residue found that when people moved from one task to another, part of their attention stayed with the first task, and their performance on the second task dropped, especially when the first task was left unfinished or was done under time pressure.2 A block with one task inside it has nothing to leave residue on.
This is also where music and timers earn their place. They are good at holding a block once it has begun. They are not good at deciding what the block is for, which is why they belong here and not at the start.
Recover: leave the screen before the next thing
Deep work shouldn’t end the second the timer rings.
Leave the screen. Move. Breathe. Reduce input for a few minutes. Give yourself a real transition instead of jumping directly from difficult work into more stimulation.
Most of what we call a break is a change of input, not a reduction of it: the feed replaces the document. Field research on micro-breaks suggests the breaks that help recovery are the ones that lower demand (a walk, a window, water) rather than the ones that add a new stream.3 The evidence is self-reported and comes from ordinary workplaces, so hold it loosely. But it agrees with what a tired brain already knows.
Return: capture where you stopped
Before you leave, write three things down. What’s finished? What’s still open? What’s the first step when you come back?
This is the cheapest step and the one almost nobody does. Unfinished tasks stay active in mind. That is the old Zeigarnik observation, and although its replications are mixed, Leroy’s residue results point the same way.4 Writing down the open loop and the first next step does two things: it lets the current session actually close, and it means the next session does not have to begin by reconstructing the last one.
That’s the difference. Deep work isn’t only about concentrating harder while you’re working. It’s about creating a repeatable way to enter, focus, recover and return.
What a whole ritual looks like in practice
A 60-minute block, end to end
Ignite (2 minutes). One sentence on paper: what will be finished. One line under it: what you are not doing. Phone in another room.
Focus (45 minutes). One tab, one task. Timer and music if they help. If something else comes to mind, write it on the same page and go back.
Recover (8 minutes). Stand up. Leave the screen. No new input: water, a window, a walk to the door and back.
Return (5 minutes). Three lines: finished, still open, first step next time. Close the tab.
The times are a suggestion. The order is not.
What this does not fix
A ritual makes starting easier and finishing cleaner. It does not make hard work easy, it does not replace sleep, and no study cited here shows that any of this “rewires” anything. If a tool tells you otherwise, ask to see the paper.
Footnotes
Sources, and their limits
- Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple plans. American Psychologist, 54(7), 493–503.Shows that 'if–then' plans about when and where to act raise the rate of actually acting. It does not show anything about the quality of the work once started.
- Leroy, S. (2009). 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), 168–181.Lab studies with students: leaving a task unfinished, or under time pressure, left more of it in mind during the next task and lowered performance on it. Effects were measured over minutes, not days.
- Zeigarnik, B. (1927). Über das Behalten von erledigten und unerledigten Handlungen. Psychologische Forschung, 9, 1–85.The original finding that interrupted tasks are remembered better than completed ones. Replications since have been mixed; treat it as a useful description, not a law.
- Bennett, A. A., Gabriel, A. S., & Calderwood, C. (2020). Examining the interplay of micro-break durations and activities for employee recovery. Journal of Occupational Health Psychology, 25(2), 126–142.Field data on short breaks during the workday: breaks that reduce input help recovery more than breaks that add it. Self-reported, one workplace sample.