Typography as Hierarchy, Not Decoration

The most common typography mistake isn't picking an ugly font. It's using type weight and size to decorate instead of to organize — reaching for bold because a phrase feels important in the moment, or bumping a line up a size because it looks better on the page, without any system behind that decision. The mistake is easy to miss because the result often looks fine in isolation. A single bolded phrase rarely looks wrong. The damage shows up later, cumulatively, once a reader has seen enough inconsistent signals to stop trusting any of them.

Bolding a phrase because it feels important, without a system behind that decision, trains the reader to stop trusting bold as a signal — because sometimes it means "important" and sometimes it just means "I liked how it looked." The first few times a reader hits an ungrounded bold, they'll pause and try to work out why it's there. After enough of those pauses turn up nothing, they learn the more efficient response: ignore bold entirely, and read every word at the same weight of attention. At that point, the tool has stopped working, for anyone, including the times it's used correctly.

This is worth taking seriously because type is doing hierarchy work whether or not that's the intent. Size, weight, and color all tell the reader what to look at first, second, and last, whether that ordering was deliberate or accidental. There's no neutral setting where type just presents information without organizing it — a paragraph with one bolded word has already told the reader that word matters more than the others around it, regardless of whether anyone meant to send that message. The only choice available is whether that hierarchy is intentional or incidental, not whether it exists.

It helps to notice how often this goes wrong not through carelessness but through good intentions applied inconsistently. A writer bolds a phrase to help a skimming reader find it later. A different writer, on a different page, bolds a phrase because it's the topic sentence. A third bolds one because it's a warning. Each individual choice is reasonable. None of them agree with each other about what bold means, and a reader moving between those pages inherits all three unstated rules at once, with no way to tell which one is in play on any given page.

The fix is treating every weight and size change as a decision with a reason attached — this is a section heading, this is an inline emphasis, this is a caption — and being able to name that reason on request. That test is more useful than it sounds, because it's easy to apply in the moment a decision gets made and hard to fake after the fact. If a designer or writer can't say, in one short phrase, what category of thing a piece of styled text is meant to be, that's a signal the styling is decorative rather than structural.

Here's what that looks like in practice, on an ordinary page of body copy. A heading gets a size and weight that means "new section starts here," applied the same way every time a new section starts. An inline emphasis gets its own distinct treatment — often italic rather than bold, precisely so it doesn't compete with headings for the same visual signal — applied only to genuinely emphasized words, not to whatever the writer personally found interesting. A caption gets a smaller size and a muted color, consistently, so the eye learns within a page or two exactly what that treatment always means.

The natural objection is that this sounds like it drains all the personality out of writing — that a system this rigid turns every document into the same beige template, and that some occasional stylistic flourish is what makes a piece of writing feel human rather than mechanical. That's a fair concern, but it conflates two different things: a considered voice in the sentences themselves, and a consistent visual grammar around them. A piece of writing can have enormous personality in its word choice and rhythm while still using bold, size, and color the same way every time it needs to signal structure.

There's also a real edge case worth acknowledging: occasional, deliberate rule-breaking, done knowingly, is different from the drift this piece is arguing against. A single pull quote rendered dramatically larger than anything around it, used once in an entire article to mark a genuinely pivotal line, isn't decoration — it's an intentional, named exception to an otherwise consistent system, and readers can tell the difference between that and a habit of styling whatever caught someone's eye that day. The problem was never variation. It was variation with no rule behind it.

Once hierarchy is consistent, readers stop having to work out what matters. That's the actual purpose of typographic hierarchy, and it's a legibility feature, not a style preference. A reader scanning a page with a disciplined hierarchy can tell, in under a second, which line is the heading, which paragraph is the lead, and which small line at the bottom is a footnote — not because they read any of it carefully, but because the visual system has already told them where each piece belongs before they've processed a single word.

The cost of getting this wrong compounds with length and complexity. A single-page flyer with inconsistent bolding is a minor annoyance; a reader can brute-force their way through a page that short. A long article, a dense report, or a product interface with dozens of screens doesn't offer that grace. Every inconsistency adds a small tax the reader pays without noticing they're paying it, and across enough content, that tax adds up to the vague sense that something is hard to follow, even when every individual sentence is perfectly clear.

This is also why hierarchy problems are undervalued in review. A misspelled word gets caught immediately, because it's a discrete, nameable error. An ungrounded bold or an inconsistent heading size rarely gets flagged the same way, because nothing about it is technically wrong — it's just one more small decision made without reference to any of the others. The fix isn't a proofreading pass. It's asking, before publishing anything, whether every styled piece of text on the page can point to a rule it's following.

None of this requires an elaborate system to get right. It requires deciding, once, what bold means, what size changes mean, and what color changes mean on a given page or in a given product — and then applying those meanings the same way every single time, even when a different choice would look more interesting in the moment. Typography that organizes rather than decorates doesn't ask readers to notice it. It asks them to trust it, and trust, once it's built, is what lets a reader stop thinking about the page and start thinking about what's actually on it.