Week 34 · Human Interaction Psychology

Who is crossing

WheelchairRunner

Wheelchair — Stopped cold — four wheels cannot climb a six-inch lip.

The Ramp Was
Never Just for One

Ronald Mace used a wheelchair for forty-six years after contracting polio at nine, and became the architect who wrote the seam in the sidewalk that almost nobody notices they are using. Flip the corner below.

Part I

Who Was It Really Built For?

Every account of the curb cut starts with the same person: a wheelchair user, stopped cold by a six-inch step. That is true, and it is also not the whole population that curb cut now serves.

Predict

Of everyone who steps through a curb cut on any given day, what share is not the wheelchair user it was built for?

Below is one real screen, not a corner. Five lenses stand in for five different ways a body or a device can meet it. The screen itself never changes underneath — only which failure it produces, and for whom, changes with the lens.

Constraint simulator

Simulated reduced acuity — a blur, not a diagnosis.

Contrast: 1.82:1; Continue button text · fails 4.5:1
Smallest target: 20px; below the 44px floor
Missing labels: 2; email field, help icon

Lens

The contrast ratio, smallest target and missing-label count in the readout above are computed from the same underlying mock in every lens — they are facts about the screen, not about the lens. What changes is the consequence: a blur makes the low contrast unreadable; the colour-vision filter makes a colour-only cue disappear; the tremor track makes the small target expensive; the reach arc puts the help icon out of range; the tree shows what a screen reader gets when nothing was labelled.

Part II

The Mismatch, Not the Person

Ronald Mace’s central move was to relocate the problem. A wheelchair user is not disabled by a wide, flat hallway; they are disabled by a staircase. The staircase is the defect, not the wheelchair.

This reframing — sometimes called the social model of disability, against an older “medical model” that locates the problem in the person’s body — is what turns accessibility from an act of charity into an engineering defect a designer is responsible for fixing. In 1997 Mace convened a working group at NC State’s Center for Universal Design that codified the idea into seven testable principles:

Equitable use — one flow serves everyone; no separate “lite” path.
Flexibility in use — keyboard, mouse, touch and voice all reach the same action.
Simple and intuitive use — predictable patterns, no jargon-only controls.
Perceptible information — state is never carried by one sense alone.
Tolerance for error — a slip should not be a catastrophe.
Low physical effort — reach and force costs compound with repetition.
Size and space for use — targets sized for the hand that is actually there.

A decade later, and independently, the W3C organized the Web Content Accessibility Guidelines around four letters: Perceivable, Operable, Understandable, Robust — POUR. The structural echo above is real; the direct citation is not. W3C’s own documentation never names Mace as its source. Read the lineage as convergent evolution, two groups arriving at the same shape from different directions, not a confirmed line from one to the other.

The design is useful and marketable to people with diverse abilities.

Ronald Mace, in the reasoning behind Principle 1

Part III

Four Fixes, Independently True

Compare the fully broken panel against the fully fixed one, then pull each of the four fixes apart to prove they are separable — not one monolithic “accessible mode” bolted on after the fact.

Isolate one fix at a time

Checks passing: 0/4

Fast parallax slide, the kind that triggers vestibular symptoms.

Four independent switches, four independent effects: turning one on never turns another off, and turning all four on reproduces exactly the “fixed” side of the comparison above — nothing left to guess at.

Part IV

Four Cases, Honestly Graded

The lineage from a poured curb to a component library runs through real institutions, real dates and at least one contested claim.

Case

1972Today

1972 — Berkeley pours the first sanctioned curb cut.

Disabled activists — the Rolling Quads, including Ed Roberts — poured concrete into a curb by hand before the city sanctioned one on Telegraph Avenue. But priority is contested: a 1940s Kalamazoo, Michigan ramp and 1930s UK pram ramps predate it. Treat "Berkeley invented the curb cut" as the popular version, not the verified first.

Secondary

Part V

Spend the Contrast Budget Yourself

Two sliders, one measured number. Push the text and the background toward each other and watch the ratio — and the two WCAG thresholds it either clears or misses — move live.

Contrast, measured live

WCAG relative luminance, computed from the two sliders below

Contrast: 1.42:1
AA text · 4.5:1: Fail; needs +3.08
AAA text · 7:1: Fail
Large text / UI · 3:1: Fail

Text lightness78%

BlackWhite

Background lightness92%

BlackWhite

Challenge · Fix the screen

Open

Four things are wrong with this sign-up panel. Tap each one below to fix it directly — solved when all four computed checks pass at once.

Contrast: 1.82:1
Smallest target: 20px
Missing labels: 2
Focus order: Out of order

Every fix is independent. Order does not matter.

Insight

Designing for the edges fixes the middle.

The curb cut was engineered for one specific mismatch and turned out to serve five. The four fixes in the panel above were each named for one specific mismatch — a hand, an eye, a screen reader, a keyboard — and none of them cost the mouse-and-monitor user anything at all. That is not a coincidence worth celebrating each time it happens; it is the mechanism, working exactly as Mace described it in 1997.

Mental model

Whenever a feature is scoped as being for one named, narrow group,

look for who else is standing at the same curb, not just the group the ticket names,

because a mismatch fixed for the edge case is a mismatch removed for everyone standing near it, at no extra cost to the people it was never blocking.

Tomorrowrun your own primary screen through the four checks above by hand — contrast, target size, labels, focus order — and count how many of the four it already passes.

End of experience

← Back to the constellation

The Ramp WasNever Just for One

Who Was It Really Built For?

The Mismatch, Not the Person

Four Fixes, Independently True

Four Cases, Honestly Graded

Spend the Contrast Budget Yourself

Where this idea goes next

The Ramp WasNever Just for One

Who Was It Really Built For?

The Mismatch, Not the Person

Four Fixes, Independently True

Four Cases, Honestly Graded

Spend the Contrast Budget Yourself

The Ramp Was
Never Just for One

The Ramp Was
Never Just for One