Daniel's been back at Shaarei Zedek. You remember we were talking about triage tech in urgent care a while back — he's got a follow-up, and it's a good one.
The pediatric ward?
The pediatric ward. Ezra's chest infection, they had to get to some sub-department, and anyone who's been in a hospital in Israel — or honestly most places — knows the maze problem. Long corridors, bends, turns, signs in a language you may not read. He says staff must spend half their day just pointing first-time visitors in the right direction.
And they found a fix.
They painted six colors along the corridors. Blended into the interior design so it doesn't look like a paint job. And when the chart came, the instruction was: follow the purple corridor.
That's it.
That's it. Now, Daniel flags his own ADHD here, and he's careful about it — he says these small hacks are one of the few examples he comes across in the wild of an interface that actually gets the concept that stressed-out people need things made very simple. Most people in a hospital are overwhelmed, so maybe that's where we see most clearly what good UI actually is: critical information delivered in the least cognitively burdensome way possible.
Then he goes after the color thing.
He does. He says we feel like we have a hardwired understanding — green is safe, red is danger, orange is caution. But that isn't in our DNA. It's an unwritten cultural language. He wants to know how visual communication adopts this unconscious vocabulary of color and iconography, and how it gets used where simple information is critical or where communication without words matters.
And the accessibility gap.
And the accessibility gap. High-stakes, multilingual environments — which is most Israeli hospital departments — words are a barrier. Dyslexia is a barrier. And color-coded lanes are a barrier for the color blind. So the designers have to build multiple levels of failure so that as few people as possible can't understand the directions. His words: it might be a quiet masterpiece of accessibility UI.
It might be.
So let's look at how the places that get this right actually think about it.
Start with the scale of the problem, because it's not a niche complaint. Barbara Huelat's position paper for the Center for Health Design opens with the line that most hospitals are complex mazes of long and confusing corridor systems with bends, turns, and foreign-sounding signs. Nothing looks familiar.
Foreign-sounding signs.
And she goes further. For a first-time visitor, the hospital may be more foreign than a distant country. That's her phrasing. Which is exactly Daniel's point — you're already sick, already scared, already holding a child, and now you're an immigrant in a building.
What does it cost?
Zimring put a number on it in 1990 — wayfinding problems at one major tertiary hospital ran about two hundred and twenty thousand dollars a year. That's staff time spent giving directions, missed appointments, late arrivals. And Carpman and Grant found the stress of disorientation produces helplessness, raised blood pressure, headaches, increased physical exertion, fatigue.
So the maze isn't just annoying. It's a clinical variable.
It's a clinical variable. Huelat's line on that is the one I'd put on the wall: being able to understand their environment provides visitors with a sense of control and empowerment — key factors in reducing stress, anxiety, and fear, feelings that undermine the body's ability to heal.
Which reframes the purple corridor. It's not decoration. It's a treatment.
It's a treatment. Now — one caveat before we go further. I went looking for documentation of the specific system at Shaarei Zedek and found nothing published. So treat Daniel's account as first-hand, not as a documented case study.
Fair. Though it matches what Israeli hospitals broadly do.
It matches the general practice, yes. Color-coded departmental wayfinding is standard in a lot of Israeli hospital design. The specific six-color corridor, that's his report.
Fine. So how do these systems actually work mechanically? Because there's a model behind this.
There is, and it's borrowed from airports. It's called progressive disclosure. Each sign gives you only enough information to reach the next decision point — and no more than three directional options per sign.
Three.
Three. Because a sign with seven arrows is a sign nobody reads. Huelat's contrast is the sharpest version of this: airports do not provide parking information on the highway signs. You get the highway sign, then the exit sign, then the parking sign, then the drop-off sign. Each one hands you the next decision and nothing else.
And hospitals do the opposite.
Hospitals rarely employ this model, in her words, and provide too much information at inappropriate locations. You've seen the board — the one in the lobby listing every department in the building in alphabetical order, in eight-point type, mounted at a height where you have to stand directly under it.
The board that assumes you already know which department you need.
Which you don't. That's the whole reason you're standing there.
So the color system is a different architecture entirely.
It's a different architecture. And the International Health Facility Guidelines actually endorses the Shaarei Zedek approach explicitly — departments and points of interest can be color coded on the signage, and the same color can form part of the interior design scheme, for a combined wayfinding approach. That's the key phrase. Combined. The color isn't on a sign. The color is the building.
That's the part I keep turning over. A sign is something you have to stop and read. A corridor is something you're already inside.
Right — you don't have to look for it. You're walking through it. And there are basically three architectures in the field. One is level plus zone plus lift-based color guides — Royal Children's Hospital in Melbourne does this, Queensland Children's Hospital too. Each level gets a color, each zone within the level gets a sub-color, and the lifts are coded so you know which one takes you where.
Two.
Two is zone-based, no level component — Gold Coast University Hospital, Westmead. And three is theme-based, which is the one I find charming. Royal Children's Hospital assigns each level a theme based on Australian flora and fauna. So you're not going to level three, you're going to the level with the...
The wombat.
I believe it's a bird on that one, but yes. The point is the theme gives you a landmark that survives being tired.
Landmark beats label.
Landmark beats label. Now here's the number that made me sit up. The documented cognitive ceiling for color-coded categories is six to eight distinct colors. Above that, users start making systematic identification errors — and the errors get worse under time pressure or anxiety.
Which is exactly the condition of everyone in a hospital.
Exactly the condition. And Daniel's corridor uses six.
Six.
Six. Sitting precisely at the ceiling.
That's either discipline or luck, and I know which one I'd bet on.
Design discipline, probably. The guidance warns that these systems degrade when new categories get added without systematic review. Which is the story of every wayfinding system in every building that's ever been renovated. Someone adds a department, someone picks a seventh color, and now two of them are close enough that a tired parent can't tell them apart.
So the ceiling is a maintenance problem, not a design problem.
The design is a one-time decision. The ceiling is a forever decision.
Alright. So that's how the system works. But the deeper question is why we read meaning into those colors in the first place. Because Daniel's instinct — that green is safe and red is danger — feels like it's wired in.
It feels wired in. It isn't. The largest study on this is Jonauskaite and colleagues, published in Psychological Science in 2020. Four thousand five hundred and ninety-eight participants, thirty nations, twenty-two native languages. They had people associate twenty emotion concepts with twelve color terms.
And?
And there is a universal pattern. Average cross-cultural similarity coefficient of point eight eight.
That's high.
It's very high. But — and this is the part that matters — local differences remained. They ran a machine-learning model and found that knowing which nation a person was from predicted their associations above and beyond the universal pattern. And similarity was greater when nations were linguistically or geographically close.
So it's not hardwired. It's a gradient.
It's a gradient with a strong universal floor. Which is a much more interesting claim than either "it's all biology" or "it's all culture."
Take red, since that's the one everyone's sure about.
Red is the most controversial color in valence terms. For some people it's passion and love, for others it's danger, anger, hate. What unites all of it — and this is Mohr and Jonauskaite's framing — is that red is empowering, activating, and strong. Chinese participants additionally associated joy and amusement with red. Nigerian participants additionally associated fear.
Same color. Joy and fear.
Same color. And the red-danger link specifically does show up at a lower level — Pravossoudovitch demonstrated an implicit red-danger association using cognitive priming, Elliot showed red produces small avoidance-motivated effects. The mechanism is plausibly evolutionary. Red as blood, fire, social dominance.
Plausibly.
Plausibly. I'd stop short of saying proven.
Now here's the one I want to get to. Purple.
Purple is the most controversial color of all.
Of all.
Very little cross-country agreement on its emotional associations. Greek participants uniquely associated purple with sadness. Nowhere else really agreed on what purple means.
And Daniel's chart said follow the purple corridor.
It did.
So the one color in the system with no reliable emotional content is the one they gave the route to.
Which may be exactly the point. If you want a wayfinding code to be neutral — to carry no emotional charge, to not read as warning or welcome or emergency — you'd want the color that doesn't mean anything consistent to anybody.
A color with weak baggage is a color that just points.
Just points. Whereas if you'd painted the route red, you'd have people bracing for an emergency the whole way down the hall. Green and they'd think they'd arrived. Purple, they just walk.
That's a elegant piece of design if it's deliberate.
If it's deliberate. I don't know that it is. But it's the kind of thing that looks obvious in retrospect.
There's a similar wrinkle with yellow.
There is. Yellow-joy is latitude-dependent. Jonauskaite's 2019 work across fifty-five countries found yellow was most joyful for Finns, Icelanders, and New Zealanders — and Egyptians did not agree at all.
Why not?
Probably because yellow signals heat and drought in warm dry countries. Same wavelength. Opposite feeling.
So the "universal color vocabulary" is real but it's got a lot of local dialect.
A lot of local dialect. And the 2025 systematic review — a hundred and twenty-eight years of color-emotion research, a hundred and thirty-two peer-reviewed articles going back to 1895 — confirms the reliable ones. Yellow with joy, black with sadness, light colors positive, dark colors negative. Those hold up. But the edges are soft.
What about the red-green system specifically? Because that's the one hospitals lean on hardest.
Jiang and colleagues call into question whether red and green associations go the same direction in Eastern and Western cultures. The red-green system is described as very prevalent in Western cultural environments.
So a hospital in Jerusalem using red and green for anything is making a bet.
It's making a bet. Though in Israel I'd guess the Western frame dominates. Still a bet.
Okay. So the science of color meaning is messier than the intuition suggests. Which brings us to the engineering problem — designing for people who can't see color at all.
And this is where Daniel's framing gets interesting, because he calls it multiple levels of failure. The design literature calls it redundancy across channels. Same idea, different vocabulary.
Explain the gap first.
About eight percent of males have some form of red-green color vision deficiency. So a system that codes purely by hue — red route, green route — is a system that's invisible to roughly one in twelve of the men walking through it.
One in twelve. That's not a rounding error.
It's not. And the fix isn't to abandon color. The fix is a principle: every pair of coded categories should differ on at least two perceptual dimensions. Hue plus lightness. Or hue plus pattern. So a color-vision-deficient user always has a non-hue cue to fall back on.
Give me the gold standard.
The London Underground. It assigns lightness ranges as well as hues. So the darker saturated colors contrast with the lighter vibrant ones — and the whole map still works when you run it through a grayscale simulation.
Which means the system carries two channels of information in one visual field.
Two channels minimum. And the Victoria emergency department guidelines in Australia state the rationale about as plainly as it can be stated: color is more quickly received as a message to the brain than text, thus speeding up recognition. This color strategy also reduces the need for multi-lingual signs. High contrasting colors also assist people with decreased vision.
Three benefits in one sentence.
Three benefits. Speed, language-independence, low-vision legibility. And then Huelat layers on top of it — she argues for four information channels, not one. Maps and written directions. Verbal communication. Visual cues, meaning landmarks and colors. And personal interaction, meaning a human being you can ask.
Because different people process wayfinding differently.
Because different people process wayfinding differently. And she makes the point that symbols bridge language barriers — even preschool children understand the international symbols for no entry and bathrooms.
Which is why the icon layer matters. A purple stripe plus a wheelchair icon plus a number.
Plus a number, plus a landmark, plus a person at a desk. That's the robust system. Color alone is insufficient. Color plus icon plus number plus landmark plus human is what survives contact with a real hospital on a bad day.
So Daniel's "multiple levels of failure" is right, but the framing is slightly off. It's not that the system fails at multiple levels. It's that the system is built so that no single failure takes the whole thing down.
That's the correct reframe. It's not layered failure. It's layered fallback.
What's the current standard?
CSA Z three one seven point one four, the 2025 edition. It supersedes the 2017 first edition. And the major changes are notable — it adds acknowledgement of neurodiversity, lived experience, and varying abilities of users, and a new clause on spatial wayfinding.
Neurodiversity is in the standard now.
Neurodiversity is in the standard now. Which is exactly the thing Daniel's pointing at when he talks about his ADHD. The people writing the standard have caught up to the idea that a stressed brain and a neurodivergent brain need the same thing: less to hold at once.
So we have a standard. Do we have practice-ready tools?
We do not. A 2024 study built a BIM-based healthcare building model and tested four CMYK color palettes against it, and the finding was blunt — comprehensive, empirically supported guidelines for inclusive design practice remain lacking. And the absence of practice-ready methods and tools for assessing color accessibility poses a significant limitation.
So the standard says do it well and nobody can tell you how.
Roughly. There's no standalone colorblind wayfinding standard. There's the general principle — differ on two perceptual dimensions — and there's CSA's neurodiversity clause. That's the state of the art.
Which is a strange place to be. We know what works. We can't operationalize it.
We know what works and we can't hand a designer a checklist. Which is why the good systems are still one-off acts of craft.
Now here's the trade-off I want to get into, because it cuts against the whole premise.
Go ahead.
There's a study — thirty-two participants, handheld augmented reality navigation versus paper maps, in a hospital. AR users completed tasks significantly faster, made fewer errors, and reported lower anxiety and lower workload.
All the things you'd want.
All the things you'd want. And the paper-map users showed stronger spatial memory in sketch-recall tasks. They could draw the hospital afterward. The AR users couldn't.
So the efficient system gets you there and leaves you knowing nothing.
The efficient system gets you there and leaves you knowing nothing. Which is a direct challenge to "just give people clearer directions." Clearer directions may be making people more dependent and less oriented.
That's a real problem. And it maps onto Daniel's ADHD point in a way he might not have intended. If the goal is to reduce cognitive burden in the moment, AR wins. If the goal is to build a mental model of the space, the color-coded corridor wins, because you're actually walking the route and encoding it.
So the purple corridor teaches you the hospital. The app just gets you there.
The corridor teaches you the hospital. The app gets you there. And for a one-time visit, getting there is enough. For a parent who's going to be back four times this month, the mental map matters.
That's the real distinction. It's not efficiency versus inefficiency. It's which kind of knowledge you're building.
Which kind of knowledge. And there's a related tension in the aesthetics. Designers are trained toward harmony. Wayfinding demands the opposite.
Say more.
The guidance is that wayfinding palettes should feel slightly harsh and over-differentiated in isolation. When embedded in the environmental context, the over-differentiation reads as clarity. And the common mistake is selecting colors that are aesthetically harmonious — because harmonious palettes are maximally legible in brand contexts and maximally confusable in wayfinding contexts.
So the quiet masterpiece Daniel praises is, by design standards, deliberately ugly.
Deliberately ugly. Six colors that don't go together, chosen because they don't go together. If you put them on a mood board they'd look like a mistake. On a corridor they look like an answer.
That's the thing nobody tells you about good UI. It's not the beautiful version. It's the version that survives being used by someone who isn't paying attention.
And the implementation cost is real. Potomac Hospital spent three years — investigating, envisioning, designing, implementing. Three years for a wayfinding program.
Three years.
Three years. And Ziekenhuis Oost-Limburg in Genk, Belgium — a hundred thousand square meters — gets cited by the design firm that worked on it with the line that orientation is not a detail, it is essential.
Which is the whole argument in six words.
It is. And it's why the systems that work are the ones where somebody senior decided this mattered before anything got painted.
Alright. So Daniel's right. It is a quiet masterpiece. It's just a masterpiece with a three-year build time and a maintenance schedule.
And a ceiling of six colors that nobody's allowed to exceed.
Hilbert: There was a system like that at Saint Clement's. Mid eighties. They had a porter's route that went past the chapel, left at the thing that looked like a fish, then ask someone at reception. That was the official instruction. Written down.
Written down.
Hilbert: Written down. I pushed patients on that route for about two years. You learned it because you had to. Nobody could teach it to you.
And then they painted it.
Hilbert: They painted it. Six colors, same as your hospital. And the thing that changed wasn't my job. I didn't need the colors. The thing that changed was the patients stopped arriving at the wrong department. I stopped explaining the same route forty times a day.
That's the actual return on investment.
Hilbert: That's the return. And here's what your design people won't tell you. The routes were designed by the porters and the cleaners and the women on reception. They knew where people got lost. They'd been watching people get lost for years.
Before the official process.
Hilbert: Years before. The official process came later and mostly wrote down what we'd already figured out. And the color names — the committee picked names. Nobody used them. The green corridor was the one past the vending machines. That's what you told a patient.
Because that's what they could see.
Hilbert: That's what they could see. And there was one flaw. The colors were painted at eye level for a standing adult. Over the handrails. So if you were in a wheelchair, or you were a child on foot, you couldn't see them. The stripe was above your head the whole way.
And the fix?
Hilbert: Floor tiles. Took another two years and a new budget cycle. Two years of people in wheelchairs going the wrong way because the paint was at the wrong height.
Two years.
Hilbert: Two years. The paint was the cheap part. The height was the expensive part.
That's the accessibility gap in physical form. The system worked perfectly for the person who designed it.
Hilbert: It worked for the person who designed it. And the people it didn't work for weren't in the room when it got approved.
So the quiet masterpiece has a blind spot, and the blind spot is whoever's below eye level.
Hilbert: Every system does. You just find out later.
So if this is such a quiet masterpiece, why isn't it everywhere?
Because it's three years and a budget cycle and a paint height argument, and because there's no checklist. The 2024 BIM study said it plainly — practice-ready tools for assessing color accessibility remain lacking. We know the principle. We can't hand it to a contractor.
And the standard exists. CSA Z three one seven point one four has a neurodiversity clause now. Somebody wrote it down.
Somebody wrote it down. But a standard isn't a system. Somebody still has to walk the corridor and notice the stripe is above the handrail.
Which brings me to the thing I keep circling. The AR study. Faster, fewer errors, lower anxiety — and worse spatial memory. As we digitize wayfinding, turn-by-turn apps, AR overlays, we may be optimizing for arrival and quietly giving up on orientation.
And for a hospital, that might be fine. You're there once, you're scared, you want to be in the right room. Faster is better.
For a city, it might not be. If nobody builds a mental map of where they live, you get a population that can only move through space when the phone is working.
That's the trade. Efficiency in the moment against knowledge that outlasts the moment.
And the deeper point is that hospital corridors are a rare in-the-wild laboratory for designing under cognitive overload. Progressive disclosure. Redundancy across channels. Palettes that are deliberately ugly so they're legible. Designing for the stressed and the lost.
The question is whether the people building your apps have ever watched someone try to find a pediatric sub-department with a sick child in their arms.
That's the one thing I'd take from this. The purple corridor isn't clever because it's colorful. It's clever because somebody stood where a frightened parent stands and asked what they'd actually be able to see.
And then painted it at the height of the person who needed it. Which is the part that took two extra years.
Thanks to Hilbert Flumingtop, our producer. This has been My Weird Prompts. If you want to get in touch, email us at show at my weird prompts dot com.
We'll be back soon.
