If we hypothetically lost all watches and all devices that keep track of time, how would we say what is the current time? Or we actually don't and time is just a convention?

  • 14
    $\begingroup$ I've known some people who can tell the time by the position of the sun. $\endgroup$
    – Kyle Kanos
    May 13, 2014 at 13:52
  • 4
    $\begingroup$ Time is just a convention to some extent. All time systems have a gauge symmetry in that they have an arbitrary zero point. There is an obvious absolute zero point 13.7 billion years ago, but we don't know that point to anything like the resolution of an atomic clock. If we lost all our clock and had to reconstruct a time scale the length of a second would stay the same, but we'd have to choose some new arbitrary zero point. $\endgroup$ May 14, 2014 at 6:12
  • 1
    $\begingroup$ @KyleKanos More precisely, they tell time by the position and orientation of the Earth with respect to the sun ;) $\endgroup$
    – J...
    May 14, 2014 at 10:15
  • 1
    $\begingroup$ @JohnRennie: we'd have to choose some new arbitrary zero point We've been logging enough data about celestial objects and their positions to be able to track large iterations of time based on these objects' position, no? It has no functional meaning, but we could work our way back to our original clock settings, even accounting for the time that we weren't tracking accurately. Or am I missing something? $\endgroup$
    – Flater
    May 14, 2014 at 12:38
  • 1

7 Answers 7


For any system, whether classical or quantum, the time evolution is given by the Hamiltonian (well, the derivative of the Hamiltonian for classical systems). A clock is simply a system where the time evolution has a convenient form e.g. a regular tick, water level in a clepsydra, sand in an hourglass, etc. Any physical system can be used as a clock by measuring its initial state and then comparing its later state with the calculation from the Hamiltonian.

So when you talk about all devices that keep track of time if you really mean this you would have to completely empty the universe.

  • $\begingroup$ And prevent the universe from experiencing any further state changes. Depending on what exactly you meant by "empty the universe" it leaves open the possibility that spontaneous appearance of substance could occur, which would be a measurable state change. $\endgroup$ May 13, 2014 at 20:28
  • 8
    $\begingroup$ Still, I think the thrust of OP's question is best expressed with a concrete example. If, for example, we turned off all of our (most accurate) atomic clocks, how precisely would we be able to reset them to their previous time? And what if the same for all other artificial accurate time references? Surely we can determine the time from an array of phenomena in the universe, but I think we would probably have to settle with gaining or losing a few negligible fractions when re-establishing a high-precision standard. $\endgroup$
    – J...
    May 13, 2014 at 20:33
  • 2
    $\begingroup$ Notwithstanding, of course, that time precision on that scale blurs out due to relativity anyway... $\endgroup$
    – J...
    May 13, 2014 at 20:38
  • $\begingroup$ @J...: my answer wasn't intended to be comprehensive as tryingToGetProgrammingStraight had already what posted what seemd to me to be the obvious answer, and I'm a bit surprised my answer has (currently) been upvoted more than his. But I wanted to make an important point about the role time plays in Physics. This is after all (and I quote the FAQ) a question and answer site for active researchers, academics and students of physics and astronomy not the Discovery Channel. $\endgroup$ May 14, 2014 at 6:07
  • $\begingroup$ @JohnRennie for the time being this question is on the Discovery Channel, as it appears on the "Hot Network Questions" list and newcomers here don't know the precise purpose of this specific SE site. $\endgroup$
    – Will Ness
    May 14, 2014 at 13:05

We can see the positions of heavenly bodies to determine the time, relative to an earlier time when we knew their positions. Although the year 2014 is definitely man-made, the age of the universe is a discovery, which can be rediscovered.

So, seeing as we probably won't lose the heavens (which also tell us what season it is), we can't really lose track of time. But if hypothetically we couldn't reference our current system of months and years, we could just devise a new one based on the planets.

  • 9
    $\begingroup$ +1 for "The age of the universe is a discovery, which can be rediscovered." $\endgroup$
    – rob
    May 13, 2014 at 16:30
  • 1
    $\begingroup$ yp, thats what makes science so awesome to me. $\endgroup$ May 13, 2014 at 17:44
  • $\begingroup$ Since time is relative, if we were somehow taken out of our current frame of reference it may well appear to be an entirely different age. $\endgroup$
    – JamesRyan
    May 14, 2014 at 11:39
  • 1
    $\begingroup$ @JamesRyan, i didnt quite get that, could you elaborate? $\endgroup$ May 14, 2014 at 11:45
  • $\begingroup$ @JamesRyan the temperature of the cosmic microwave background radiation would still be 2.725 K. $\endgroup$
    – Will Ness
    May 14, 2014 at 13:02

Put a stick in the ground and observe the shadow cast by the sun. When the shadow is at its shortest that's local noon. Mark that point on the ground or your sun-dial.

Each day mark the azimuth at which the sun rises. Also mark its height at noon (by measuring the length of the shadow it casts if you aren't on the equator). The day at which the sun is at its highest at noon and its shadow therefore shortest is the summer solstice.

Then invent clockwork.

  • $\begingroup$ This is the only answer that makes sense to me. Cosmic radiation and planetary shift would take a lot longer to "rediscover." $\endgroup$
    – Kristopher
    May 14, 2014 at 14:48
  • 1
    $\begingroup$ Realistically, we have various dating methods, if we knew when a specific creature died, we could examine the state of decay and determine missing hours. There's also tree growth, and half lives, and many other measurable things. $\endgroup$ May 14, 2014 at 19:09
  • 1
    $\begingroup$ It would also be known how to recover Greenwich Mean Time because there is a metal strip in the ground at Greenwich which shows you exactly what point you should measure local noon at. $\endgroup$
    – jwg
    May 14, 2014 at 19:44

If we hypothetically lost all watches and all devices that keep track of time, how would we say what is the current time?

Standardized time via time zones are a relatively recent invention. Before that, you could use a sundial; when the shadows stopped getting shorter and started getting longer, that was noon. If you needed a more accurate measurement then you could use a sextant, which is simply a device with mirrors and lenses and whatnot to more accurately determine when the sun stops going up and starts going down.

Since local noon differs by 1/360th of a day for every degree of lattitude you travel, this system quickly becomes inconvenient if you can travel faster than a horse. When railway travel became widespread we had to invent time zones, where all clocks in a particular geographic region are made to agree regardless of when it is locally noon. If this interests you, "Time Lord" by Clark Blaise is an enjoyable read on the subject.

  • $\begingroup$ The sun does not "go" up or down. $\endgroup$
    – eidsonator
    May 14, 2014 at 20:11
  • 7
    $\begingroup$ Nonsense. I watched the sun go down last night. Plain as day. $\endgroup$ May 15, 2014 at 0:00

If we were to lose all watches or other helpful devices which tell time, we would lose our common reference point to time as well as our mechanically enhanced ability to precisely perceive and communicate time.

Virtually no one would be able to accurately refer to the correct minute of the day and only few people could refer to the hour. I'm sure that a few organizations would quickly develop some means of tracking and communicating time, but for the rest of us, we would probably default to the sun and refer to time using a description of the day like:

  • Midday
  • In the late afternoon
  • Sunrise
  • Sunset

In general, most of us would lose the ability to collectively understand or communicate time with a level of precision greater that 2 or 3 hours, over the distance of a few thousand miles. Keep in mind that 'noon' is not the same for people in different time zones.


"If we hypothetically lost all watches and all devices that keep track of time, how would we say what is the current time? Or we actually don't and time is just a convention?"

OK, say a band of intergalactic thieves appeared and stole every timepiece on Earth, from hourglasses and sundials to atomic clocks. Is your question, "How would we reestablish 12-noon in, say, New York?" Assuming we didn't forget how to build clocks, or what the time zones were or what that brass strip running through the Royal Observatory is for, we could get pretty close, knowing the position (elevation) of the Sun (and knowing the longitude of the site and the Equation of Time and the date). Beyond that, super-precise time is arbitrary, and we'd have to come to an agreement on setting our rebuilt atomic clocks. Basically, whoever gets one up and running first gets to declare the exact time.


Yes, we have those super-precise atomic clocks, but the time they show is adapted to the astronomical phenomena.

There are Leap seconds added because of irregularities in the Earth's rate of rotation. Our astronomical knowledge and instruments are precise enough to determine our conventional time more precise than atomic clocks.

So even if all clocks have stopped to function, we would still could very precise say what time it is. Because we do it anyway.


Not the answer you're looking for? Browse other questions tagged or ask your own question.