how are warning labels designed?

Design leaders often say (or used to say? I don’t know anymore) that good design is “invisible” — we notice it mostly when it’s bad. But for warnings, the goal is opposite of invisibility.

I’ve done a lot of digging in the past — from related grad school courses, freelance work, and just my own noodlings — and it turns out the the look of warning labels evolved from a long history of semiotics, design-led standardization, and scientific research.

Without really thinking much about it (I assume), you probably realize that warnings are mostly pictures. Text is secondary. (People read in different languages, and anyway, nobody’s going to stop and read an essay before sticking their hand in machinery.) The pictures and symbols used in warnings range in their realism — from depicting an actual thing, like a person holding a handrail, to communicating abstract meanings, like radioactivity.

Some symbols have pre-existing meaning and established use. For example, people have been drawing stick figures since prehistoric cave paintings. A circle plus a few lines doesn’t look much like any of your friends, but it’s a quick shorthand that we all agree about. When someone designs a warning label, they don’t need to totally make up a symbol to represent a human, and they don’t need to draw a realistic-looking figure either.

Another familiar symbol, a circle with a line through it to represent “no” or “don’t do this thing,” had history prior to its use in modern warnings. It’s similar in form to the Greek letter Theta, which was used as a shorthand to indicate death. For example, Theta shows up in the famous Gladiator Mosaic to mark the gladiators who died in combat.

Even with some symbols becoming intuitive over time and use, the urgency of standardization arose in the early 1900s alongside the growing popularity of automobiles. (Imagine the chaos of cars intermixed with horse-drawn carriages and pedestrians, with no reliable signage to anticipate a sharp curve or railroad crossing.) In 1909, a Paris convention on International Circulation of Motor Vehicles developed some standardized road signs — at the time, they were all white pictures on blue backgrounds. By 1939, the League of Nations had further designated 26 uses of color and shape in standard road signs, including the ubiquitous red octagon stop sign.

Over the next bunch of decades, many design trends converged on simplicity and understandability of visual communication. There were also increasingly better-defined standards for safety and accessibility.

In 1974, the United States Department of Transportation asked the American Institute of Graphic Arts (AIGA) to develop a standardized set of pictograms to use in signage for public spaces like airports. The committee compiled an inventory of existing symbol systems, gave them scores, and then combined and redesigned them into a single unified set. These pictograms often use a “flat man,” a simple figure with consistent weights and shapes who also shows up in warning labels. With these established symbols (and more added over time), it became easier for people find, recognize, and interpret signs everywhere they went.

For communicating safety-critical warnings, Human Factors researchers in the latter part of the 20th century established rules that overlap with other work on symbols and signage, but with more rigorous testing. These warning-design rules are based on how people notice, perceive, and respond to information. For example:

• Warnings need to draw attention — so besides the visual design, the warning’s size and location in the environment are relevant.

• Layouts are relatively consistent, using a signal word, pictogram, and direct wording to explain the hazard and how to avoid it. For signal words: DANGER is the highest risk, followed by WARNING, and then CAUTION. Red indicates the highest level of hazard, followed by orange, and then yellow.

• Pictograms are simple, to be understandable at a glance. Think two-dimensional, solid outlines.

• If a hazard is too complicated for one image, some warning labels provide multiple pictograms to communicate cause and effect — almost like a un-amusing comic strip.

• The warning label as a whole needs to be widely understandable. Labels are redesigned and retested until statistically up to snuff. (In my experience doing this kind of work, it’s almost shocking how often a proposed symbol can actually be “critically confusing” — sometimes accidentally communicating the opposite of its intent.)

Okay, was this boring? I think it might have boring. I cut out a lot of extra fluff and detail to make it less boring.

miscellany

• Pottery Barn is sayin’ this Super Mario collab is for teens but I’d like to own over half these items, especially the glow-in-the-dark pajamas and Super Star string lights.

• I started playing Hollow Knight in the spooky season and realized that it may actually be a long-haul game, based on how slowly I progress past bosses. I foresee myself giving up at some point, but the hand-drawn art style and ambiance are top notch. (There’s a sequel coming, but at my current rate I’ll get to it in 2030.)

• Johnson & Johnson updated their classic scripty logo (based on the founder’s signature) for something unquestionably boring. I mean, “easier to read.” I’m all for readability but I don’t think that applies to logos. Another one bites the dust.

• The Beatles gifted us a new song, half from the grave. It’s fine. But it’s an interesting creative and technological output, having started as a Lennon demo, then an unfinished 1990s project by the other three, and finally some AI work made possible by Peter Jackson’s technological advances with the Get Back documentary. I feel like John would have something quippy and quotable to say about it, but I don’t, sorry.