Slow looking: a practical routine that works on any object

Why eight seconds isn't enough

Eight seconds buys you the headline: subject, size, dominant colour, a mood, a guess. Everything that actually distinguishes the work — the second figure, the colour that shouldn't be there, the detail that changes the story — lives past the point where most people move on.

The Smithsonian's National Museum of Asian Art, which runs guided slow-looking activities, puts the average glance at under thirty seconds and treats the practice as learnable in an afternoon — its published sessions start at three to five minutes per work and build toward fifteen or twenty. The entry cost really is that low: the skill is staying, not expertise.

What the extra time buys is better evidence. First impressions are data, but only the first layer — description done properly makes interpretation possible, which is the same reason this site's readings separate what is visible from what is concluded.

There is also a quieter payoff: the eight-second glance is where confident mistakes breed. The longer you stay, the more the object gets to correct your first account of it — which is why every museum that teaches looking teaches duration first.

The three-minute version

For a crowded room or a first pass. A quiet phone timer helps — Tate recommends ten minutes as an ideal, but notes that even two deliberate minutes beat the eight-second average.

1. 01

   Minute one: look only. No label, no phone, no conclusions. Let your eyes go where they go, and notice the first thing that keeps pulling them back.

2. 02

   Minute two: describe silently. In your own words, say what is literally there — shapes, colours, marks, materials — as if describing it to someone who cannot see it.

3. 03

   Minute three: ask one question. Form a single question the work doesn't answer — why this colour, what happens next — and spend the last minute looking for whatever evidence it points at.

The fifteen-minute version

For a work that earns a chair. MoMA's slow-looking itinerary and the Smithsonian's guided activities use roughly this shape; treat the timings as guides, not rules.

Fifteen minutes is a long time to stand in one place, so claim the conditions first: a bench, a quiet corner, a work that isn't carrying a crowd at its back. MoMA's version starts at five minutes and builds gradually — here, the timer marks the exercise's edge, not a deadline to fill.

1. 01

   Settle (2 minutes). Find a comfortable position with a clear view — sit if you can — and simply look, without trying to accomplish anything.

2. 02

   Describe (3 minutes). Work through the whole surface in your head or under your breath: what is literally depicted or built, edge to edge.

3. 03

   Change your position (2 minutes). Stand, step left, get close enough to see the marks, then far enough to lose them. Note what changes and what doesn't.

4. 04

   Follow one detail (3 minutes). Pick the element that most holds your attention and trace everything it touches — colour echoes, repeated shapes, what it does to the space around it.

5. 05

   Notice yourself (2 minutes). Pay attention to your own responses — impatience, calm, a memory surfacing — as data about the looking rather than distractions from it.

6. 06

   Leave a summary (3 minutes). Before you walk away, say or write three sentences: what it is, what surprised you, and what you would look at first next time.

What to do when nothing new appears

Every routine hits the wall — minute four, apparently nothing left. Tate's own list of entry points is the fix: texture, colour, shape, symbols, story, perspective. Pick one and look only for that.

The museum playbooks add a few restart moves. Look at the edges and corners, where painters put what they didn't want noticed. Count a thing — windows, hands, reds. Or hunt what is missing: the shadow that should be there, the door that isn't.

And boredom itself is worth one honest minute. Sometimes a work is doing nothing on purpose, and noticing that is noticing something.

Slow looking with ordinary objects

The routine is object-agnostic — the Smithsonian guide frames slow looking as something you can practise on anything, and designed objects are arguably ideal practice. Every detail on a teapot, a trainer or a biscuit tin was decided by someone, which means the question 'why is it like this' always has an answer worth looking for.

Pick something you own that you have never properly looked at — the wear pattern on a tool handle, the lettering on a packet — and run the three-minute version on it. The noticing transfers; that is the entire point of the practice.

A note on where this pays off beyond museums: the same routine turns a junk-shop shelf into a much better hunting ground. Objects you would have passed at eight seconds announce their repairs, reprints and honest wear at three minutes — which is also, not coincidentally, the difference the rest of these guides keep circling.

After looking: let a reading test what you saw

Slow looking is the ideal preparation for a scan: the reading it returns is AI-generated and provisional, so its observations mean more when you already have your own to hold them against. Photograph the work well, run the reading, and put its 'Look closer' list beside your notes. Agreement is satisfying; disagreement is the interesting part.

Where the reading names things you never saw — a symbol, a technique, a period detail — that is the next loop of looking: go back to the object and find them. Where the reading is plainly wrong, that is a find too, and the five-minute audit tells you how wrong and where.

Either way, the looking comes first. A reading works best as a second pair of eyes, not a first pair.

None of this requires a museum. The fifteen-minute version works on a poster in a corridor, a building on your street, the back of a camera — anywhere a thing has more in it than a glance will hand you.