What happens just outside your focus, and why a word choice can move real behavior

1. Most of what you’re looking at right now, you aren’t actually seeing

Look at the ringed dot below and hold your gaze. Don’t look directly at the others; just notice what happens to them.

They flicker and seem to vanish. Peripheral vision doesn’t blur what’s outside your focus; it compresses it into a rough statistical average, pooling nearby shapes together rather than storing each one. MIT researcher Ruth Rosenholtz demonstrated this by running Boston’s 2013 subway map redesign through a model of that compression: the old map dissolved into noise, while the new one held its shape.

The evolutionary reason is practical: central vision is built from densely packed cone cells for detail, while peripheral vision is rod-dominated and built to catch motion. For an animal watching for predators, that tradeoff was survival.

2. The ‘harder to read helps you remember’ idea is more nuanced than it first appears

In 1917, critic Viktor Shklovsky argued that familiarity dulls perception: friction forces attention. A 2010 study by Connor Diemand-Yauman tried to test a version of that idea: people remembered more from a hard-to-read font than an easy one. It became popular design advice, including in this book.

Later research made the picture less clear. A 2018 meta-analysis and several direct replications did not find a consistent memory benefit from illegible fonts.

Typography had already made this bet once before, with mixed results. Sans serif letterforms were called ‘grotesque’ when they appeared in the 1830s, built for signage and headlines rather than body text. Their reputation as the more legible, modern choice came later, driven largely by Bauhaus-era ideology rather than the actual legibility research of the time, which mostly found serif and sans serif performed comparably for extended reading. The pattern is familiar: a clean story about perception and clarity gets adopted faster than anyone tests it.

What does hold up, for learning at least, is a narrower idea: desirable difficulties such as spaced practice or self-testing. This is effort that makes you retrieve information, not effort that just makes a page unpleasant to look at.

3. Calling someone a voter moves more people to vote than asking them to

Researchers Christopher Bryan and Gregory Walton tested two versions of a pre-election survey: one asking how important it was to vote, the other how important it was to be a voter. Using real state voting records, the noun version produced turnout 11 percentage points higher. They repeated it in a different state and found the same result.

The real-world version of this is ‘Don’t Mess with Texas.’ In 1985, Texas was spending $20 million a year cleaning highway litter, most of it dumped by young men who tuned out standard anti-littering appeals. Instead of telling that group to stop littering, the campaign reframed littering as un-Texan, an identity violation instead of a rule violation. Litter dropped 29% in the first year and roughly 72% within five years, a result tracked and confirmed for decades since.

It is the same mechanism at a much larger scale: framing a behavior as who you are can move people further than framing it as a rule to follow. It shows up in product copy too, with ‘become a member’ versus ‘sign up,’ usually without anyone naming why it works.

The pattern

Each idea became more useful once it was placed in context. That is the finding worth keeping: a compelling design principle is a starting point, and following the evidence helps reveal where it applies, where it needs narrowing, and what it can genuinely teach us.