What We Owe Each Other's Attention: The Small Ethics of Interruption

There is a gesture small children make that I have not been able to shake. She takes my face in both hands and turns it, physically, toward her. No words. Just a correction of where I was pointed.

What lands is not the interruption. It is the assumption underneath it — that my attention is a thing with a direction, that the direction matters, and that pointing it somewhere else while she was speaking was a kind of absence she had every right to object to.

She is right, and she has not yet learned the adult convention of pretending otherwise. Among grown people we have agreed that attention is free to take. Someone's time we treat as theirs; you ask before booking an hour of it. Their body is obviously theirs. Their consent, in most domains, is a thing we know we need. But their attention we lift dozens of times a day without asking and without noticing, and we have built an entire economy on the assumption that doing so at scale is a business model rather than a moral question.

Why Attention Belongs Alongside Time and Consent

The philosophical case for taking attention seriously is older than the technology that made it urgent.

William James, writing in 1890, put it about as bluntly as it can be put: my experience is what I agree to attend to. That is not a metaphor about focus. It is a claim about what a life is made of. The hours you lived are not the hours you were physically present for; they are the ones you attended to. Everything else happened near you.

If that is true, then attention is not a resource like electricity, interchangeable and replaceable. It is closer to the raw material of a life. Simone Weil went further and treated attention as a moral capacity — the thing that makes it possible to see another person as they actually are rather than as a shape in your peripheral vision. Iris Murdoch built much of her moral philosophy on the same move: that the ethical work happens in the looking, long before any decision gets made. By the time you are choosing how to act toward someone, the important part is already over. You have decided what you are willing to see.

Herbert Simon noticed the economic consequence back in 1971, in a line that has aged uncomfortably well: an abundance of information creates a scarcity of attention. He was describing an institutional problem. He was also, accidentally, writing the business plan for the next fifty years.

Put those together and the conclusion is not complicated. If attention is what a life is made of, and it is genuinely scarce, then taking someone's without their agreement is taking something of theirs. Not metaphorically. The only reason it does not feel like theft is that the amounts are small and the norms are absent.

Every Interruption Is a Small Ethical Choice

This is the part that reframes an ordinary day.

Interruptions get treated as neutral events, like weather. They are not events. Each one is a decision someone made about the relative value of their convenience and your continuity. Most of the time the decision is made without any awareness of being a decision at all, which is precisely what makes it worth naming.

Consider what actually happens when you send a message marked urgent that is not urgent. You have made a claim on someone else's mental state. You have said, in effect: what I want to hand off right now is worth more than whatever you were holding. Sometimes that is true. A genuine emergency earns it. But most of what gets sent that way is not an emergency — it is a transfer of discomfort. Holding an unresolved thing is unpleasant, and passing it along makes it briefly less unpleasant for you.

The costs are not evenly distributed either. Research on interrupted work keeps arriving at the same uncomfortable place: the recovery is far longer than the interruption. A question that takes thirty seconds to ask does not cost thirty seconds. It costs the thirty seconds plus the several minutes of reassembly, plus the residue — the part of your mind that stays partly on the previous task and does not fully arrive at the next one. The person interrupting pays none of this. They experience a resolved item. The full price is charged to someone who did not agree to it.

I want to be careful here, because there is a version of this argument that curdles into something unpleasant. The point is not that other people should leave you alone so you can be productive. That is the same instrumental logic as the attention economy, just with you as the beneficiary. Being interruptible is part of belonging to people. A household where no one can ask a question is not a household with good boundaries; it is a household where people have stopped mattering to each other.

The distinction is not interruption versus no interruption. It is whether the interruption is a considered act or a reflex. A friend who calls because something has gone wrong is not violating anything. A notification engineered by a team of people to arrive when you are most likely to be idle is a different category of act entirely, and calling both of them "getting a message" hides the thing that matters.

What Changes When the Taking Is Industrialised

Here is where I have to be honest about my own position. I write software. I have built interfaces, and building an interface means making decisions about what pulls the eye. Everyone in this profession has, at some point, made something more noticeable than it strictly needed to be. The attention economy is not a distant villain. It is a set of ordinary professional choices, made by people who mostly thought of themselves as improving engagement metrics.

What scale changes is not the ethics of the individual act. It is the accumulation. One well-designed prompt to return to an app is a small thing, easily forgiven. Four hundred of them a day, tuned by measurement against millions of people, is not four hundred small things. It is a system with a stake in your inability to sustain a thought — because a sustained thought is time not spent in the feed.

Matthew Crawford's argument in The World Beyond Your Head is that we have quietly permitted the enclosure of a commons. Silence used to be the default condition of public space; now it is a premium product, sold back to you as a quiet carriage or a paid tier without ads. We would not accept a company piping advertisements into a public park at a volume that made conversation difficult. We accepted the same thing in the informational space we all live in, mostly because it happened one reasonable-seeming increment at a time.

James Williams, who worked in advertising technology before writing about it, made the observation that stays with me longest: the concern is not that these systems distract us from what we want to do. It is that sustained exposure changes what we want to do. A distraction you can resist is an inconvenience. A system that gradually reshapes your preferences is something else — it interferes with the part of you that does the choosing.

And the corporate defence has always been consent. You agreed. You installed it, you accepted the terms, you can leave whenever you like. But consent given once, to a system that continuously optimises against the person who gave it, is doing very little work. We understand this in every other domain. We do not consider a contract binding if one party rewrites it daily in response to measurements of the other party's weaknesses.

Norms for a Household

The industrial version is not mine to fix. The interpersonal version is, and it is where most of my attention is actually taken and given anyway. A few norms that have survived contact with real life.

Ask before you spend it. "Do you have a minute?" is not a formality. It is a genuine request for consent, and it needs to be askable in a way that permits a no. A household where the answer is always yes has just relocated the obligation.

Batch what can be batched. Four questions delivered at once cost a fraction of four questions delivered across an afternoon. This is a small discipline with a large effect, and it is mostly a matter of writing things down instead of transmitting them the moment they occur to you.

Name the priority when you interrupt. "This can wait until tonight" and "this is now" are enormously different pieces of information, and supplying it costs the sender nothing. Most of the harm in casual interruption comes from the receiver having to evaluate urgency themselves, which requires fully loading the thing you sent.

Protect the transitions, not just the work. The most valuable protected time in a house with children is rarely the middle of a task. It is the first twenty minutes after someone walks in the door, and the last hour before sleep. Those are the windows where the day's real conversations either happen or don't.

Put the phone somewhere with a door between you. Not as an act of discipline. As an act of removing a competitor from the room. The device does not need to be looked at to take something; the mere possibility of it arriving occupies a corner of the mind.

Give the full thing when you give it. Half-attention is worse than a delayed yes. "Give me five minutes and I'm entirely yours" is more honest and more generous than nodding along with a screen in your peripheral vision, and children in particular can tell the difference with an accuracy that should embarrass us.

Norms for a Workplace

The same principle, translated into a place where the incentives run the other way.

Default to asynchronous; make synchronous the exception you justify. A meeting is a request for several people's attention simultaneously, which is the most expensive thing you can ask for. If it could have been a written update, the cost was incurred for the convenience of the organiser.

Treat "urgent" as a limited currency. Everyone knows who in an organisation marks everything urgent, and everyone has learned to discount them. That is what depleting a shared resource looks like at the level of a team.

Publish when you are reachable rather than being reachable always. Two guaranteed windows a day beats being nominally available for nine hours and actually available for none of them. Predictability is the thing colleagues need; instant response is a poor substitute for it.

Do not schedule over the only unbroken block someone has. A calendar with a two-hour gap is not empty. It is the only place the difficult thinking can happen, and a meeting placed in the middle of it does not take thirty minutes — it takes the whole block.

Say what the meeting needs from each person. Attending something you cannot contribute to is attention taken and nothing produced. If someone is there in case they are needed, say so, and let them leave when they are not.

Notice who pays. The people whose attention gets taken most casually are usually the ones with the least standing to object. That asymmetry is the clearest sign that the norm is not really a norm but a hierarchy in disguise.

Questions Worth Answering Honestly

Isn't this just an elaborate justification for being unavailable?
It can become that, and the failure mode is worth watching for. The claim is not that your attention is too precious for other people. It is closer to the opposite — that it is valuable enough to give deliberately rather than leak continuously. Someone who protects their focus all day and then gives none of it to the people they live with has not understood the argument.

Where does this leave phones in social settings?
The practical finding I trust most is that a visible phone degrades a conversation even when nobody touches it, because both people are aware that a third party could arrive at any moment. That does not require a ban. It requires putting it face down, out of sight, or in another room — and being willing to say why.

Is any of this actually the individual's problem, given the scale of the industry?
Both are true at once, and the industrial version genuinely does need regulation and design change rather than personal virtue. But the interpersonal version is not a smaller version of the same problem — it is a different problem, and it is the one that determines whether the people around you feel seen. That one is entirely yours.

What about people whose work requires constant availability?
Then the norm applies to their employer rather than to them. Someone on call is having their attention purchased, and the honest question is whether the compensation reflects what is actually being bought — not just the hours, but the impossibility of being fully anywhere during them.

How do I raise this without sounding precious about it?
Describe the mechanism, not the principle. "I lose about twenty minutes rebuilding after a question, so can we batch them to after lunch?" travels much better than a claim about the sanctity of focus. People respond well to specifics and badly to being told they are doing something ethically suspect.

I sit for Heartfulness meditation in the mornings, and the honest report from most sessions is not tranquillity. It is watching how relentlessly the mind volunteers itself elsewhere. Twenty minutes with nothing arriving, and it will still find something to be somewhere else about. What that practice has actually taught me is not how to concentrate. It is how much effort attention costs, which means I now know roughly what I am asking for when I ask for someone's — and roughly what I am taking when I don't ask.