LLBBL Blog

LLBBL Blog

Day 27: People Have Tried This Before

Yesterday I dug into the astronomical anchor problem: whether civil time should stay tied to the sun, and the complications that come with loosening that tie. Redesigning civil time at scale is extremely complex, and I don’t pretend to have the answer. But people have already tried, more than once, to do something about it. Today I want to look at those attempts and what we can learn from how each one landed.

Believe it or not, the urge to rationalize and unify time is much older than the technology that finally made it possible. A continuous atomic timescale has existed for a while now, but people were trying to redesign civil time well over a century before that.

The thread running through all of it: none of these attempts failed for technical reasons. The hard part of changing civil time has never been the engineering. It’s getting people to agree.

So, let’s take a look at the first attempt.

Decimal time, French Revolution, 1793

The French Revolution, as part of its broader push to metricize everything alongside the meter and the gram, legislated a decimal time system. A day became 10 hours, an hour became 100 minutes, a minute became 100 seconds. That’s 100,000 decimal seconds per day, each about 0.864 standard seconds.

It was elegant. It coupled cleanly with the rest of the metric system. The math was easier. It was decreed in 1793 but formally mandatory for only about six months, from September 1794 to April 1795, and almost universally ignored. Watchmakers kept making 24-hour watches. People kept reading them. The government quietly stopped enforcing it and abolished it in 1795.

The lesson, which is going to repeat: top-down imposition of timekeeping reform fails without overwhelming, immediate, individually-felt benefit. The metric system worked because shopkeepers and farmers and engineers felt the benefit at every transaction. Decimal time didn’t change anyone’s life enough to overcome the friction of relearning every clock.

Swatch Internet Time, 1998

Two centuries later the Swiss watch company Swatch tried something with a similar spirit. They invented Swatch Internet Time, a zone-free format for the internet era. A day split into 1000 “.beat” units of 86.4 seconds each, the same instant everywhere on Earth. @500 meant the same moment in Tokyo, São Paulo, and New York. The anchor was Biel Mean Time, UTC+1, the time zone of Swatch’s headquarters.

It actually got adopted, briefly. The Dreamcast RPG Phantasy Star Online used .beats for scheduling international play because it sidestepped the time-zone mess of coordinating Japanese, American, and European players.

It failed because of the anchor. Tying global time to “Swatch’s HQ” guaranteed it would read as a marketing stunt. You can’t get the world to adopt a time standard whose name advertises a watch company. But the lesson here is the inverse of decimal time: bottom-up, voluntary, internet-native adoption of a zoneless format is possible. The format was usable. People used it. To grow beyond a niche it just needed neutral, institutional backing, something more like ISO than a brand.

TAI and GPS time are already working

A continuous, universal, zoneless time scale has existed for decades, and several billion people use it every day without knowing it.

GPS time started ticking on January 6, 1980, with no leap seconds. It’s currently 18 seconds ahead of UTC, and the gap grows every time UTC adds a leap second. It’s the time satellite navigation runs on, and the relativistic corrections that keep it honest are what I dug into back on Day 11. Every position fix your phone gets reads GPS time. Aviation, financial exchanges, mobile networks, they all sync to it.

International Atomic Time (TAI) is older and less famous. It began on January 1, 1958, is currently 37 seconds ahead of UTC, and is maintained by the BIPM in Paris as a weighted average of around 400 atomic clocks, each counting the cesium second we unpacked on Day 8. It has ticked without interruption and has never had a leap second. If you wanted a universal civil time tomorrow, TAI is worth considering.

What we have now is just a UTC-shaped facade. The civil layer has kept the appearance of solar anchoring. The engineering layer underneath went atomic decades ago. We already have a hybrid system.

Hanke-Henry Universal Time, 2012

We covered the Hanke-Henry calendar on Day 24, but the proposal had a second part that gets less press, and it’s worth being honest about how thin it is. Alongside the calendar, Hanke and Henry advocated abolishing time zones and putting the whole world on UTC, so everyone shares one clock. A meeting at “14:00” is at 14:00 everywhere. In Tokyo the sun is descending, in Los Angeles it’s the middle of the night, in Lagos it’s morning. The clock is the same. The sky is not.

That’s about as far as it goes. They worked out the calendar in detail, but the universal-time half is closer to a one-line aspiration: adopt UTC, drop the zones. There’s no transition plan and no account of how billions of people switch without chaos, which is telling, because that is the hard part. For all that, the proposal got real coverage, in Smithsonian, The Independent, and on CBS News, and it picked up a bit of political interest along the way. But none of that supplied the missing adoption plan, so “just use UTC” still hasn’t gone anywhere. We can’t ignore that barrier: the adoption plan is the whole problem.

Single-timezone experiments

A few countries have run partial versions at national scale. Since 1949, all of mainland China uses UTC+8, even though it spans five geographic time zones. In Xinjiang the official sunrise is around 10 AM in winter, and an unofficial “Xinjiang Time” runs two hours behind even though it isn’t legally recognized. India puts the whole subcontinent on IST (UTC+5:30). Russia keeps merging and splitting its zones for political and administrative reasons.

The lesson is that a single-zone time works, with some friction at the edges. People can live an hour or two off solar local, as long as their daily activities line up with their employers and neighbors. “Noon must equal solar noon” turns out to be a culturally negotiable constraint, not a biological one.

What all of this suggests

Five lessons from the prior attempts:

  1. The engineering problem is solved. TAI, GPS time, the atomic second. We have continuous universal timescales that have run for decades.
  2. Top-down forced reform fails. Decimal time, the French Republican Calendar. Mandated by powerful states, ignored, rolled back within years.
  3. Branded reforms fail. Swatch Internet Time was reasonable but tied to a corporate identity that delegitimized it.
  4. Single-anchor wide zones work in practice. China and India have done it for decades.
  5. The Hanke-Henry universal-time proposal The barrier is coordination, not technical.

If I had to draw one generalization, it’s that the path to actually changing civil time probably isn’t a sudden global switch or a top-down mandate.

It looks more like the IANA tz database: an open, voluntary, technically credible standard that earns adoption gradually, one institution and one country at a time, with a clear migration path from where we are now.

Tomorrow I want to close the loop. We already toured how strange time gets the moment you leave Earth, back on Day 12. Tomorrow’s question is the harder one: knowing all that, what would a clock for an interplanetary civilization actually have to be? The people designing it right now are converging on an answer, and it looks a lot like the reform Earth keeps refusing to make for itself.

Sources

I’d appreciate a follow. You can subscribe with your email below. The emails go out once a week, or you can find me on Mastodon at @[email protected].

History Time 30daysoftime Standards Calendars