For thousands of years, we've divided days into 24 hours, hours into 60 minutes, and minutes into 60 seconds. But why do we have to do that? Here's the story of the one gloriously failed attempt to decimalize time.
Top image: Steve Wilde/Flickr.
It's easy to forget just how arbitrary the way we measure time — or measure anything, for that matter — really is. Yes, the Earth's rotation means it makes sense to look at one spin around the planet's axis as a unit, so that gives us the day. And the time it takes the Earth to revolve once around the Sun makes sense as another unit, so that's the year. The month is not quite as intuitive as those first two, though its etymology from the word "moon" is a rather gigantic clue. It is indeed derived from the lunar cycle, and there's evidence that humans have marked the monthly changes of the Moon as part of timekeeping since the Paleolithic.
But weeks, hours, minutes, and seconds? There's no compelling reason why we must divide them the way that we do, other than the fact that that's how we've apparently always done it. Weeks have been anywhere from three to ten days long. In a rather pleasing inversion of what we're used to, the Sinhalese people in ancient Sri Lanka divided the day into 60 Peya, which we might consider their version of the hour, and then in turn divided these into 24 Vinadi. This means a Vinadi is exactly as long as a minute — just arrived at in the opposite way.
In general, it does seem that most ancient cultures settled on dividing the day into twelve parts, or first splitting this into day and night halves and then dividing by twelve. The exact reason why so many civilizations settled on multiples of 12 or 24 to divide the day is lost to history. The fact that there are about twelve lunar cycles in a year might have something to do with it, or it might also have been because it's really easy to work with the number twelve. Our current system of 24 hours divided into 60 minutes, which in turn is split into 60 seconds each, comes from the Babylonians, who themselves used a base-60 system. Speaking of which...
The Base Problem
The fact that we build our number system around ten is an evolutionary fluke. Certainly, it's a very ancient evolutionary fluke — five-fingered hands go right back to some of the very first organisms to emerge from the ocean. The two hands are our most convenient counting aids, and it's only natural that our species would be drawn to organizing our numbers in terms of this fact of our anatomy. We let computers have base-2 in their binary codes, but otherwise base-10 is so omnipresent that it's difficult for us to even contemplate alternatives.
That's a shame, because honestly, base-10 is really pretty crap, mathematically speaking. Because the divisors of ten are just 2 and 5, fractions involving multiples of 3 and 4 are necessarily clunky. People who like to argue about the best possible base for our number system, quite apart from the fact that they're clearly the life of any party, tend to point to base-12 as the clearly preferable option. There's certainly something to this - instead of .333..., 1/3 would be .4 in base-12, and 1/4 gets shortened from .25 to .3. 1/5 has to become the repeating decimal .24972497..., and 1/7 is still a mess, but it's arguably still a worthwhile trade-off to get simpler decimals - or duodecimals, I should say - for the thirds and fourths.
While humanity has settled on base-10 for counting, the usefulness of base-12 endures in our measurements. Beyond the fact that we divide the day into two sets of 12 hours apiece, "dozen" and "gross" still persist as special terms for 12 and 122. There's twelve inches in a foot, and the same used to be true of the relationship between ounces and pounds before the old Troy system was replaced by Avoirdupois, which uses a 16-ounce pound.
Time to Decimalize
Throughout history, not all have been comfortable with the fact that we split time between base-10 and base-12, and so various pushes have been made for us to abandon one in favor of the other. While there's a theoretical argument that it would be better to go completely duodecimal, the pervasiveness of base-10 meant that was never really an option. The only way forward was redefine all our units of measurements in units of ten.
This, of course, is the motivation behind the metric system. The idea was first proposed back in 1586 by a Flemish mathematician and engineer named Simon Stevin, who argued for the widespread adoption of decimal notation, with the use of decimal fractions presaging the inevitable introduction of decimal weights and measures. Robert Hooke's mentor (and my non-ancestor) John Wilkins advocated for what would be the metric system when he proposed the idea to the Royal Society in 1668, and the ensuing decade saw a great deal of writing on the virtues of decimalization.