——

title: ‘Why a small red circle makes you check your hair’

date: draft: true summary: ‘The recording dot on your screen inherited its authority from neon gas, a 1962 semiconductor, and a Japanese engineer who refused to quit. Nobody designed what it does to you. It accumulated.’ description: ‘The recording dot on your screen inherited its authority from neon gas, a 1962 semiconductor, and a Japanese engineer who refused to quit. Nobody designed what it does to you. It accumulated.’ tags: [design, ux, conventions, history, physics] category: design

It is Wednesday afternoon. The energy that carried the morning is gone.

I am in the sound booth at the office – glass walls, a loud voice contained. This is Research and Design, the weekly session where the full design community gathers, synchronous and otherwise. It is the kind of call that gets recorded automatically, the moment the first person joins. Before I connect, I check my hair. It can get out of hand. The red dot is already there when I arrive. I notice it the way you notice a person standing very still in a room. Not alarming. Just present. Watching.

My back straightens anyway.

The call is easy for me today. The team owns the content – I am here for information, for amusement, for the occasional chat comment that lets me react without interrupting. Camera on. Microphone muted. A face in the grid tells the presenter someone is paying attention. The red dot does something similar, but to everyone. Posture adjusts. Faces compose. The version of yourself you would rather not broadcast quietly steps aside.

Nobody decided this would happen. That small circle never earned its authority. It inherited it – through a chain of physics, accidents, and engineering decisions that stretches back nearly a century, to a time before anyone imagined video calls or recording software or the particular social anxiety of knowing someone might watch this later.

——

What neon does when you apply voltage

The story begins in the 1930s, in the indicator lamps wired into electrical panels, radio sets, and industrial machinery.

These were NE-2 neon glow lamps – small glass capsules containing neon gas at low pressure, with two electrodes inside. When you applied voltage, electrons collided with neon atoms, exciting them to a higher energy state. As those atoms returned to their ground state, they released that energy as photons. The wavelength of those photons – the color of the emitted light – is determined by the quantum energy levels of neon itself. You cannot change it by adjusting a dial. It is a property of the atom.

Neon atoms emit at around 640 nanometers. That is orange-red. Not because anyone chose it. Because that is what neon does.¹

The NE-2 was cheap, durable, and could run directly from mains voltage without a transformer. It lasted 25,000 to 100,000 hours. For decades, it was the default answer to the question “how do we show this device is powered on?” And the answer was always the same color, because physics left no room for preference.

Red became the visual language of “this thing is working” – not by design, but by the emission spectrum of a noble gas.

——

The semiconductor that could only make one color

By the 1960s, engineers knew that solid-state electronics were replacing vacuum tubes and gas lamps in many applications. The question was whether a semiconductor could emit light efficiently enough to replace the NE-2.

In 1962, Nick Holonyak Jr., a consulting scientist at General Electric’s laboratory in Syracuse, New York, demonstrated the first practical visible-light LED.² He was working with gallium arsenide phosphide – GaAsP – a compound semiconductor in the III-V family. Where most of his colleagues were focused on infrared-emitting devices, Holonyak was deliberately working in the visible spectrum. “I wanted to work in the visible spectrum where the human eye sees, and everybody else was working in the infrared,” he later said.³

The reason his LED was red is the same reason neon indicators were orange-red: physics.

An LED emits light when electrons and holes recombine at a semiconductor junction, releasing energy as photons. The wavelength of those photons – the color – is determined by the bandgap of the material: the energy difference between the electron’s initial and final states. GaAsP’s bandgap corresponds to photon energies in the red visible range, around 650 nanometers.⁴ Holonyak did not choose red. He chose the material that worked – the one whose crystal structure he could grow reliably, dope controllably, and coax into emitting light rather than heat. That material happened to have a bandgap that landed in the red.

Getting to other colors meant finding semiconductor compounds with different bandgaps. Green requires a wider bandgap. Blue requires a wider one still. And wider-bandgap materials are, as a rule, harder to synthesize, more difficult to grow as crystals without defects, and more demanding in their doping requirements. Each color is not an adjustment. It is a separate materials science problem.

Holonyak understood this. He called it the “alloy road” – the idea that by tuning the composition of III-V semiconductor alloys, you could in principle produce LEDs of any color.⁵ It was a clear-eyed vision of where the technology could go. But vision and capability are different things. In 1962, the capability ended at red.

——

Ten years to yellow

The first step along the alloy road took a decade, and it came from an unexpected direction.

George Craford had been one of Holonyak’s doctoral students at the University of Illinois. In 1967, he joined Monsanto Chemical Company, which was at the time one of the largest LED manufacturers in the world. Craford’s team was trying to push the emission spectrum of GaAsP toward shorter wavelengths – toward yellow, then green. The direct approach was not working. Moving the composition of GaAsP toward more phosphorus shifted the emission toward green, but the efficiency collapsed, producing devices too dim to be useful.

The breakthrough came from an unlikely source: a seminar by a Bell Labs researcher who mentioned the use of nitrogen as a dopant in gallium phosphide LEDs.⁶ Bell Labs had suggested that nitrogen doping did not improve GaAsP devices. Craford disagreed. He believed the Bell Labs experiments had been done with inferior crystal quality, and that his team could grow better material. “We decided that we could grow better crystal and the experiment would work for us,” he later recalled.⁷

They were right. By adding nitrogen to GaAsP, Craford’s team created devices that could emit throughout a range from red-amber to green, with efficiencies more than an order of magnitude better than previous attempts.⁸ In 1972 – ten years after Holonyak’s first red LED – Craford demonstrated the first yellow LED.⁹

When the result was announced, the reaction from industry was not what Craford had hoped. “The initial reaction was, ‘Wow, that’s great, but our customers are very happy with red LEDs. Who needs other colors?’”¹⁰

The customers had been trained by neon. Red was what an indicator was supposed to look like.

——

The problem that nearly everyone abandoned

Yellow was progress. But the color that mattered most – the one that would complete the spectrum and make white LED light possible – was blue. And blue was a genuinely hard problem.

Blue light has a shorter wavelength than red or green, around 450–470 nanometers. To emit at that wavelength, you need a semiconductor with a wider bandgap than GaAsP could provide. The candidate material that most researchers agreed was theoretically promising was gallium nitride – GaN. It had the right bandgap. It had been known since the 1960s. The problem was making it work.

Specifically, the problem was p-type doping – introducing the kind of electron vacancies (“holes”) that you need to create a functional p-n junction. GaN was stubbornly resistant to it. Researchers at RCA had tried in the late 1960s and given up. By the 1980s, the consensus in the field was that GaN was a dead end.¹¹

Then, at Nichia Corporation – a small chemical company in Tokushima, Japan, known primarily for phosphors – a young engineer named Shuji Nakamura decided to try anyway.

Nakamura had joined Nichia in 1979 straight from university. In 1988 he spent a year at the University of Florida, where he learned the latest techniques in compound semiconductor growth. When he returned to Japan, he wanted to work on GaN. Nichia did not have the budget for a state-of-the-art growth system. What they had was an older machine – but rather than accept the limitation, Nakamura spent months rebuilding and modifying it himself, improving its temperature control and gas flow until it was capable of producing higher-quality GaN layers than most well-funded labs could achieve.¹²

He then threatened his supervisors that he would resign if they did not give him resources to continue the GaN research.¹³ They agreed.

In 1993, after years of work on crystal growth, doping techniques, and device structure, Nakamura produced a bright, efficient blue LED using indium gallium nitride – InGaN.¹⁴ The light output was 125 microwatts at 440 nanometers. It was, by the standards of what had come before, extraordinary. The Nobel Committee later described the blue LED as “a rare example of a technology that had an immediate and clear benefit to society.”¹⁵

In 2014, Nakamura, together with Isamu Akasaki and Hiroshi Amano – who had made essential contributions to GaN crystal quality and p-type doping – was awarded the Nobel Prize in Physics.¹⁶

For his work at Nichia, Nakamura was paid his regular salary and a special bonus of ¥20,000 – approximately 200 US dollars.¹⁷

——

Ninety years of unbroken inheritance

Let us return to the Wednesday afternoon sound booth.

By the time Nakamura’s blue LED arrived in 1993, the red indicator lamp had been the default visual signal for “this device is operating” for sixty years. It had moved from neon gas to solid-state electronics without anyone pausing to reconsider whether red was the right choice. Each generation of engineers had copied the previous one, because the convention looked familiar, and familiarity felt like correctness.

The red recording dot on your video call software is a direct descendent of that neon glow. It traveled from the NE-2 indicator lamp through Holonyak’s GaAsP junction to consumer electronics to broadcast cameras to recording decks to the interface conventions of digital software. No one at any point in that chain sat down and decided: red should mean “you are being observed, compose yourself accordingly.”

It accumulated.

Studio engineers wired red lamps to record-enable circuits because red was what lamps were. Broadcast camera operators learned to watch the red tally light because studios already used red for “live.” Consumer camcorders inherited the convention from professional equipment. Software designers inherited it from camcorders. And somewhere in the middle of that chain, the color stopped being a neutral indicator and became something that straightens spines and prompts hair checks.

That is worth sitting with. A convention that governs the behavior of millions of people in meetings, classrooms, and recorded presentations – a convention with real social weight and real physiological effect – was not designed. It was physics, then inertia, then familiarity, then assumption. At no point did anyone evaluate whether red was the right signal for “this is being recorded.” It was already there.

——

The question this leaves

I do not think this is only a story about indicator lamps.

The same mechanism – physics or circumstance creating a convention, inertia preserving it, familiarity making it feel inevitable – operates throughout the interfaces we build and inherit. The assumption that a shopping cart icon means “items selected for purchase.” The assumption that a hamburger menu contains navigation. The assumption that progress is shown left-to-right. Each of these was set by a specific context, at a specific moment, for reasons that made sense then and may or may not make sense now.

Most of them have never been re-examined. They were already there.

The red dot has been shaping human behavior since the 1930s. It was never designed to. It just kept getting copied. Every convention you design around was decided by someone – except most of them weren’t. They accumulated. Physics, inertia, familiarity, repetition – until the arbitrary became invisible and the invisible became correct.

What are you copying right now that nobody designed either?

——