Red is not danger everywhere

The same wavelength, different meanings

Colour association is learned by repeated pairing inside a culture, so it varies exactly as much as cultures do.

• White is weddings and purity in much of Europe and North America, and mourning at funerals across much of East Asia and parts of South Asia.
• Red is danger, debt and stop in Western finance and signage, and luck, celebration and marriage across China and India. A financial chart in red says loss in London and something closer to good in Shanghai, which is why some Chinese exchanges use red for rising prices.
• Green carries religious significance across the Islamic world, and is money in the United States because of the currency.
• Purple is mourning in Brazil and Thailand, and royalty in Britain, for the mundane reason that Tyrian purple dye was ruinously expensive.

None of this comes from the light. It comes from what the reader has repeatedly seen a colour attached to. That is good news, because it means you can find out rather than guess: look at what the actual audience’s own institutions, packaging and festivals do.

What is not cultural

A few colour effects are about the eye and hold broadly:

• Saturated colours attract attention more than desaturated ones, which is why one accent works and five do not.
• Pure blue is a poor colour for small text. The eye focuses different wavelengths at slightly different distances (chromatic aberration) and the cones sensitive to short wavelengths are sparse and nearly absent from the very centre of the fovea. Small saturated blue type on a bright background reads as slightly soft no matter how good the screen is.
• Pure red on pure blue vibrates unpleasantly, because the two focus at opposite ends of that error and the eye cannot settle on both.
• Around one in twelve men of northern European descent has some form of red-green colour vision deficiency, which the next lesson takes up properly.

These are constraints rather than meanings, and they do not vary by market.

The rule that survives every culture

Never carry meaning in hue alone.

A status dot that is green for working and red for broken conveys nothing to a substantial minority of your readers, nothing in greyscale print, and nothing on a badly calibrated projector. Add a second channel every time: a tick and a cross, a filled and hollow shape, a word, a position.

The cheap check: view the design in greyscale. Every phone and operating system has a greyscale accessibility setting, and any image editor will desaturate a screenshot. If the information survives, the colour was decorating; if it disappears, the colour was load-bearing and needs a partner.

Colour psychology claims are weaker than the marketing pages say

You will read that blue increases trust, that orange drives purchases, that red creates urgency. These claims trace back to a small number of studies with small samples, inconsistent replication, and effect sizes nowhere near what is claimed. The much-repeated “red button outperformed green” result is a single test on one site with one context and one set of surrounding colours, from which the general law does not follow.

What holds up better, and is unexciting: contrast against the surroundings predicts noticing far more reliably than hue. A red button on a red-heavy page is invisible; a green one on the same page is conspicuous, and the reverse holds on a green page. The effect people attribute to the colour is mostly the effect of the difference.

A palette you can actually work with

Define colour by role, not by name, and keep the accent scarce:

• page: background
• ink: body text
• muted: secondary text
• line: borders and rules
• accent: exactly one thing per screen

Then check each pairing for contrast before you fall in love with it. Free tools: the WebAIM contrast checker in a browser, the Colour Contrast Analyser as a desktop app, GIMP or Krita for desaturating a mock-up, and any browser’s developer tools for inspecting a live page.

The limitation, stated honestly

Cultural colour associations are generalisations about very large groups of people, and inside any of those groups they vary by age, religion, region, class and context. Red at a wedding and red on a warning label are different reds to the same person. A designer applying a country-level colour rule to an individual audience is making the same mistake as a designer applying an American rule everywhere; the granularity changed, the guessing did not.

The defensible version is narrow: know that the associations differ so you stop assuming yours are universal, then look at what your actual audience’s own materials do, and reserve your confidence for the physiological constraints and the greyscale test, which hold everywhere.