The sentence Hemingway left unfinished
Ernest Hemingway had a working rule he came back to again and again in his letters and in A Moveable Feast: he stopped writing each day not when he ran dry, but when things were going well and he still knew what came next. He would end mid-scene, sometimes mid-sentence, leaving a thread hanging that he could pick up the following morning. It sounds like a small superstition. It is actually one of the most practical things anyone has ever said about starting work.
Most of us do the exact opposite. We push a task until we hit a wall — until we are tired, stuck, or the thing is finally, blessedly finished — and then we stop. Which means the next time we sit down, we are staring either at a blank beginning or at the same wall that beat us yesterday. The hardest part of almost any work session is the first ten minutes, and we keep engineering our stopping points to make those ten minutes as hard as they can possibly be.
There is a name for the force Hemingway was quietly borrowing, and it comes from a psychology lab in Berlin nearly a century ago.
The Ovsiankina effect
In the late 1920s, a psychologist named Maria Ovsiankina ran a deceptively simple experiment. She gave people small tasks — puzzles, models, manual jobs — and interrupted them partway through, on some pretext. Then she created a quiet opportunity for them to do something else, and watched what they actually did.
The overwhelming majority went back. Without being asked, often without quite registering why, they returned to the interrupted task and finished it. Some resumed a task they had been told was already over. The pull to complete an unfinished action was strong enough to override the instruction to move on.
This is the Ovsiankina effect: an interrupted task generates a tension that seeks its own resolution. You have probably met its more famous sibling, the Zeigarnik effect — the way an unfinished task keeps intruding on your thoughts while you are trying to do something else. Zeigarnik described the mental half: it stays on your mind. Ovsiankina described the behavioral half: you get up and go back to it.
Why an open loop pulls harder than a fresh page
Both findings came out of the same circle of researchers around Kurt Lewin, and Lewin gave them a shared explanation. When you form an intention to do something, he argued, you create a kind of psychological tension — a quasi-need — that persists until the action is carried out. Finishing the task discharges the tension. Leaving it undone keeps the system charged.
That reframes the whole problem of starting. A cold task is hard to begin partly because it carries no momentum: there is no tension pulling you toward it, only the friction of a standing start pushing back. But a task you left deliberately open already has a charge on it. When you sit down, you are not starting. You are resuming. And resuming is a fundamentally different psychological act — it runs downhill, not up.
This is why writers talk about the terror of the blank page and almost never about the terror of the half-written one. A half-written page is arguing with you. A blank page is indifferent, and indifference is much harder to answer.
The strategic interruption
So the move — call it the Hemingway bridge — is to stop on purpose before you are done, at a point where you know exactly what comes next.
The counterintuitive part is the timing. Your instinct is to stop at a low point: when you are stuck, drained, or bored. That is the worst possible place to leave off, because tomorrow you inherit the stuckness with none of the warmth you had built up. Instead, stop at a high point — when the next sentence is obvious, when you can see the next three steps of the problem, when you are almost sorry to quit. That reluctance is exactly the tension you want to bank overnight.
You are not being lazy by leaving work undone. You are pre-loading tomorrow's momentum.
How to leave yourself a bridge
A few concrete ways to do it. Stop mid-motion, not at a clean boundary — mid-sentence, mid-function, mid-paragraph, halfway through the step you already know how to finish. The rougher the edge, the easier it is to grab.
Before you walk away, leave one line for your future self: not a to-do list, just the single next move. "Next: rewrite the second paragraph so it opens on the objection." One breadcrumb, placed where your eyes will land first.
And resist the urge to do the one last thing. Finishing that last small piece feels satisfying precisely because it discharges the tension — which is the opposite of what you want. Let it stay charged. The discomfort of walking away mid-thought is the price of a frictionless return.
When you should finish instead
This is a technique, not a religion, and it has a clear boundary. Small, self-contained tasks should be closed, not bridged — a two-minute reply left open just becomes one more nagging Zeigarnik loop, mental clutter with no upside. The bridge is for long, continuous work you will come back to: writing, coding, studying, designing, anything you cannot swallow in one sitting.
There is also a limit on quantity. Open loops carry a real cognitive cost; a dozen half-finished threads do not energize you, they fray you. The skill is to leave one deliberate bridge into your most important work, not to scatter unfinished business across your whole life and call it a strategy.
Building the bridge into your day
The honest difficulty is that stopping well requires a stopping point you did not choose in the heat of the moment — because in the moment, momentum tells you to either grind to exhaustion or quit at the first snag. A timer solves this. When something external ends the session, you are far more likely to stop at a natural high point rather than push into the wall.
This is where a focus timer becomes more than a countdown. In Tally, a Pomodoro session ends while you are still mid-task — and that interruption, the one most timers treat as a nuisance, is the strategic interruption Ovsiankina described. Because Tally also lets you stack that focus block onto an existing daily cue, the return is doubly easy: a familiar trigger brings you to the desk, and an open loop is already waiting there to pull you in. You spend less of your day starting from cold and more of it simply picking up where you left off.
If your hardest ten minutes are always the first ten, it is worth trying the opposite of what instinct suggests: stop before you are done, and let tomorrow start itself. You can build that rhythm with Tally.