Category: Geek stuff

_{In keeping with the dictates of thesis writing, and the sage comments of those who suggest that blog entries are not the best use of time, I resolve the following: posts on this blog between now and the completion of a draft thesis shall be limited to no more than one substantive and one narrative post per day, the latter to generally include a photograph. Posts that pertain directly to the substantive content of the thesis, as designated by the M.Phil Thesis category, are exempt from these restrictions.}

An article in Harper’s that I first took up because of its hyperbolic title – “Clash of the Time Lords: Who will own the measure of our days?” – is actually a really interesting demonstration of how human beings try to make the world fit within our understanding.

In particular, the article hinges on the fact that the second has two distinct definitions. The first is based on astronomical phenomena: 1/86,400th of a day, that being 1/365.25th of the time it takes for the Earth to orbit the sun. The second is based on the extremely precise oscillation of cesium atoms, the measuring stick used in atomic clocks. Specifically, it is 9,192,631,770 oscillations. The trouble arises from how those two are not the same; the Earth does not sweep through its orbit with perfect precision. Rather, it wobbles, hits things, and slows down. As such, astronomical time ‘slows down’ as compared with atomic time.

Right now, this is corrected for using occasional leap seconds. Every time the Earth has lagged behind atomic time by one second, one second is added to the reckoning of atomic clocks. Since our orbit continues to slow, leap seconds need to be inserted with ever greater frequency. This is good for astronomers, since it lets them continue to aim their telescopes in the same way as before. What is more controversial is whether this is a sensible system overall.

From a galactic or universal perspective, it doesn’t seem too reasonable. Ultimately, it is a throwback to the era when it was believed that the Earth occupies some metaphysically special place in the universe. When we concede that it is just one of uncountable numbers of things zipping about under the influence of gravity and other forces, the idea that time should be altered to correct for the peculiarities of its orbit becomes a difficult one to maintain, for any reasons aside from the practical ones of astronomers. Consider, for instance, the question of whether it would be appropriate to subtract a period of time if a comet or asteroid impact cause the orbit of the Earth to speed up.

That said, there are a good number of practical reasons to consider fiddling with time to match our orbit. The disjoint between calendar time and astronomical time is the reason for the piecemeal and difficult shift the world has made from the Julian to Gregorian calendars. Indeed, the point was to return key astronomical events, like the equinoxes, to the points in the calendar where they ‘should’ be. That shift famously required the negation of eleven days. For those who followed the decree of Pope Gregory XIII, they were October 5-14th, 1582. People in the UK and US didn’t switch systems until later, erasing September 3-13th, 1752. As such, the measure of time differed be eleven days across the English channel. When the US and UK did make the switch, the passage of calendar time was re-aligned with the experience of astronomical time.

Over thousands of years, allowing atomic time to rule, and diverge to an ever greater extent from astronomical time, would shift the seasons into ever different positions within the calendar. Very slowly, sunrise and sunset times would get out of sync with the times of day when they previously happened: likewise, the solstices and equinoxes. Already, the GPS satellites, which rely critically on super precise time, are 14 seconds ahead of UTC (Coordinated Universal Time). This is because they have not been counting leap seconds. Time in Unix computer systems also ignores leap seconds. As more leap seconds are added to UTC, that gap will grow.

Perhaps this is the most obvious solution: acknowledge the split and come up with two separate accountings of time: one that just counts the oscillations of those cesium atoms and thus the actual number of atomically defined seconds that pass, and another that corrects those figures for the peculiarities of our passage through space. Most people would probably only bother with the latter, but having the former as a kind of absolute record of how much time has passed since event X strikes me as more honest.

PS. Slightly related to the above is this excellent comic about dinosaurs planning to steal the prototype kilogram (the actual hunk of platinum–iridium that defines the unit of mass).