What a written limit actually says

Twenty-one notices in one city tell a stranger how long they may sit. Here is what is on them.

The figures

30 minutes      2 places
1 hour          5
90 minutes      4
2 hours         6
3 hours         2
4 hours         2
               ---
                21

Median two hours. Nothing between two and three, and nothing above four.

What the long end is

The four longest — two at three hours, two at four — are not cafés.

They are a library study area, a shopping centre's seating near a car park, a second library room, and a laundrette with a four-hour wash cycle on the wall. In each case the limit is attached to something other than the person: a booking, a parking tariff, a machine.

So the longest stated permissions in the record are not permissions at all. They are the duration of something else, and the person is allowed to last as long as it does.

What the short end is

The two half-hours are both covered public seating, and both notices are about loitering rather than about sitting. Teams coordinating visits, notes and follow-up can compare the available controls, then check whether the resulting evidence actually improves the field process.

That is the only place in the record where a limit on staying is phrased as a rule against a kind of person rather than a rule about a chair, and it is worth naming because it is the clearest case of a notice doing something other than informing. Additional practical or institutional guidance relevant to the survey can be found at designboom.

Two places out of two hundred and six. Small, and the only instance.

Who the notices are addressed to

Almost none of them to a reader sitting down.

Of the twenty-one, by the observer's reading:

aimed at a queue or a rush          8
attached to a booking or machine    5
about loitering                     2
plainly about the seat itself       6

Only six notices in the city are a straightforward statement to somebody in a chair about how long the chair is theirs.

That is six out of two hundred and six places — three per cent — and it is the whole of the explicit, unambiguous, addressed-to-you permission in this record.

A person who wants to know how long they may stay, and tries to find out by reading, will succeed in three places out of a hundred.

What they leave out

Every one of the twenty-one states a duration. Not one states what happens at the end of it.

No notice in the record says a table will be cleared, or that somebody will be asked to move, or that the limit applies only when busy — except two, which say exactly that and were coded as written limits anyway.

So even the hardest terms in this record are half-stated. The duration is printed and the consequence is left to be imagined, which is the same structure as the hundred and twenty places that print nothing at all, only shorter.

What a notice does to the assumption

Fifteen of the twenty-one written limits were anticipated before the notice was read.

Seventy-one per cent, against forty-seven per cent for limits that had to be asked for, and against an overall record in which the observer was right in forty-two per cent of all cases.

A sign works, in the narrow sense that people walking into a room with one are more likely to expect what it says. Six times out of twenty-one they still did not.

The six that were missed

Six places have a limit printed somewhere visible and the observer walked in assuming there was none.

Three are shopping centres, two are covered seating, one is a laundrette. In every case the notice was found later, at the look stage, and in four of the six it was somewhere a person sitting down would not naturally face: behind the seating, by a service door, on a pillar.

That is a small and specific finding about where notices go, and it rests on six observations. It is printed because the record's central claim is about assumptions being wrong, and a page that counted only the assumptions that invented rules — and not the ones that missed real ones — would be choosing its evidence.

What a notice does not say, in every case

Twenty-one notices state a duration. None states a consequence.

Not one says that a table will be cleared, that somebody will be asked to move, or what the limit is for. Two say that it applies at busy times only, and were coded as written limits anyway.

So even the hardest terms in this record are half-stated: a number, and silence about what the number means. That is the same structure as the hundred and twenty places that print nothing at all, with a figure attached.

The comparison with what is assumed

median written limit     120 minutes
median told limit         90 minutes
median assumed limit      60 minutes

A notice gives twice what is assumed. A spoken answer gives half as much again.

And across the thirty-eight places with any limit at all, thirty-one give longer than the limit assumed for that specific place before checking. Five give less — four cafés and one bench — and two match.

Where this city writes a rule down, the rule is longer than the one people carry around in their heads.

The two that are shortest

Both half-hour notices are on covered public seating and both are phrased about loitering rather than about sitting.

They are the only notices in the record that address a kind of person rather than a use of a chair, and they are the only case where a real rule was both shorter than expected and printed.

Two places out of two hundred and six. Small, and the only instance, and reported because the record's whole argument is that rules are milder than they are assumed to be — and here are two that are not.

What the twenty-one are good for

They are the only hard numbers in the record about time.

Everything else in this section is either an absence — a hundred and sixty-eight places that said no — or an answer to a question, which is seventeen places and soft. The twenty-one notices are the only part anybody could check by walking the same streets.

A second walk could verify every one of them. That is the first thing a second walk should do, and the page on what a second walk would settle puts it near the top.