Notification Triage Fatigue

A simple explanation

A notification arrives. Before you have decided whether it is important, another one arrives. You glance, you classify, you dismiss — and somewhere underneath that small act, an even smaller act of fatigue accumulates. By midday the fatigue has a name in your body even if not in your mind: a faint reluctance to look at the screen at all, paired with a stronger reluctance to look away.

This is not about volume alone. It is about being asked, dozens of times an hour, to act as your own editorial desk — to decide what is urgent, what can wait, what to read, what to mute, what to feel. The Reward System has framed the sorting itself as productivity. The body keeps a different ledger.

An everyday example

You sit down to write. Before you finish a sentence, a banner slides in: a colleague reacted to a message. You glance — the reaction was a thumbs-up to something three days old. You return to the sentence. A second banner: a calendar reminder for a meeting that has already been moved. A third: a news alert that you read in the bathroom and have already half-forgotten.

You write the next sentence. Each banner took less than two seconds of attention. The hour, however, produced two paragraphs instead of a page. By evening you cannot recall a single notification that mattered, but you can recall a low background ache behind the eyes. The day was not interrupted by anything large. It was interrupted by many small classifications that the system was not built to perform at scale.

Why does sorting messages feel like work?

Because it is work. Each micro-decision draws on the same finite resource that complex thinking draws on. The Reward System treats notification triage as a reward-shaped activity — there might be something good, there might be something due, there might be something status-related — and so it keeps the sort running even when the input is overwhelmingly trivial.

The exhaustion is not the volume of information. It is the cumulative cost of asking is this important? hundreds of times a day with no real way to know in advance. Triage without criteria is the most expensive form of attention, and most modern notification streams arrive without criteria attached.

The behavioral loop

A loop that hides because each step looks free:

1. Buzz — an alert arrives, often pre-empting whatever was previously in attention.
2. Pre-classification body cost — before the content is registered, a small autonomic adjustment occurs: a brief orienting response, a faint cortisol tilt.
3. Glance — the alert is glanced at, often within two seconds of arrival.
4. Micro-decision — urgent / not urgent / read later / dismiss / open. The decision is made on insufficient information.
5. System credit — the Reward System logs the sort as productive vigilance; a faint felt-event of I am on top of this arrives.
6. Return cost — attention returns to the prior task carrying a residue of the alert's content, even if the alert was dismissed.
7. Compounding — the next alert arrives before the residue has cleared; the sort runs again, on a system already mid-recovery.
8. End-of-day verdict — the body registers fatigue without an obvious source. The mind concludes it was busy.

Emotional drivers

Four feelings that keep the triage running:

• A faint anticipatory anxiety about missing something whose absence would be costly later.
• A reward-shaped curiosity about what each alert might contain.
• A status-adjacent worry about being slow to respond.
• A low-grade dread about looking at the inbox, paired with an equal dread about not looking.

What your nervous system does

The orienting response — a quick autonomic adjustment toward a novel stimulus — was evolved for occasional, high-signal interruptions. When the orienting response is triggered every two to three minutes, the system stops returning fully to baseline between alerts. Heart rate variability falls. The vagal brake stays partially disengaged. The body settles into a low-grade vigilant tone that looks, on the outside, like ordinary working posture.

Over weeks and months, the system adapts by dulling the response. The orienting becomes shallower, the sort becomes more reflexive, the felt-cost per alert appears to drop. The cumulative cost, however, does not drop — it is now paid in a flatter, more diffuse exhaustion that the loop-runner often attributes to age, sleep, or workload.

The DojoWell interpretation

Notification triage fatigue is a clean instance of effort without deposit. The Reward System's original ask was vigilance — knowing what matters, being responsive to what is asked of us, not missing the call from the person who needs us. The substitute it supplies is the felt-event of staying on top of it: the small rush of clearing a banner, the faint sense of being current, the social-shaped reward of fast response. The sort feels like work because it is work, but the work produces almost nothing.

The contacted vigilance — knowing what genuinely needs response and giving it that response — leaves a deposit. The substituted triage leaves only a thinned sense of which alerts ever mattered. The effort is real, the residue is real, the deposit is near-zero, and the density verdict is low not because notifications are bad but because the system was never built to classify them at the rate they arrive.

This is also why the closure_pattern is substituted rather than absent. Something does close at each glance — a small loop of did I miss anything? gets a small no. The closure is real. The need it was satisfying — to be in honest contact with what matters — was not.

How do I stop reacting to every alert?

You do not stop the alerts from arriving. You stop being the sorting layer for streams that should have been sorted at the source. The Reward System's framing of triage as productivity is what is workable.

Three moves, in order of difficulty:

1. Audit one day of alerts. For twenty-four hours, write down which alerts actually required action and which did not. The ratio is almost always startling, and the audit converts an abstract complaint into a usable list.
2. Move the sort upstream. Turn off banner notifications for the categories that the audit revealed as low-yield. Let those streams collect; check them on your terms.
3. Set a single arrival window. Designate a time of day when arrivals are read; outside that window, the phone is allowed to accumulate. The accumulation rarely costs what the System predicts.

Practical steps

1. Turn off badges for everything that is not a person. App badges weaponise the orienting response on streams that did not earn it.
2. Allow lock-screen previews only for messages from specific contacts. The default is everything; the workable default is a short list.
3. Move email off the phone entirely for one week. The week is the experiment, not the policy. The data the body returns will tell you what to keep.
4. Replace checking-on-arrival with checking-on-schedule for one stream. Start with the cheapest one. The System's prediction that the sky will fall is almost always wrong.
5. End the day with a short review rather than a long sweep. Five minutes of intentional inbox triage produces more deposit than four hours of micro-glances.

Reflection questions

• Which alert stream, if turned off entirely for a week, would cost you almost nothing?
• When in the day is your triage cheapest, and when is it most expensive?
• Whose actual messages do you most fear missing, and is your notification setup currently optimised for them?
• Where has triage fatigue begun to show up as a dulled response to alerts that genuinely deserved one?

Frequently Asked Questions