To measure time, we use clocks. From sophisticated atomic clocks to everyday wall clocks. When we measure the length of a time interval using a clock, we are really measuring the number of times the clock completes a cycle of events which we assume is periodic. My question is pertaining to how we know a chain of events is periodic. How do we know that a clock, say an atomic clock, is perfectly periodic? All we can do is measure the interval between its repititious cycles with another clock, not knowing whether the second clock is taking the same time to repeat its 'ticking'. Can we ever be truly sure that if some events are being repeated, the time interval required for every repetition is the same? What evidence may we provide for the same? Clocks only measure number of repetitions. Can we be sure they measure time? What could time even mean?

  • 3
    $\begingroup$ I recommend moving this question to the Philosophy stack exchange. In physics, clocks measure time partially because we define them to, and partially because in every physical way we have ever come up with, their measurements are consistent with measuring time. Philosophers, however, do indeed grapple with the question of "what does time mean," and have come up with several alternative ways to view time which may be right up your alley. $\endgroup$
    – Cort Ammon
    Mar 31, 2019 at 23:30
  • 2
    $\begingroup$ Since different clocks, built in different ways, keep being synchronized (although there may be small random fluctuations), we accept the easiest explanation that is we are witnessing the same thing repeating periodically. $\endgroup$ Apr 1, 2019 at 0:09
  • $\begingroup$ See arxiv.org/abs/0708.2743 and arxiv.org/abs/0805.4452 $\endgroup$ Apr 1, 2019 at 1:27

2 Answers 2


Clocks work according to the laws of physics. For example a caesium atomic clock uses the energy difference between two energy levels of the caesium atom to generate the tick.

Using the laws of physics we can calculate the time evolution of the clock and we find it has a regular tick. You can argue about exactly what we mean by time but as far as we are concerned it is a coordinate like any other coordinate, and calculating the evolution of the system along this coordinate is generally straightforward.

For the clock not to tick at regular time intervals one of two things would have to be true:

  1. the time experienced by the clock isn't the time as the time experienced by the observer reading the time from the clock

  2. the laws of physics change with time

Actually it's fairly easy to make option (1) apply as we can change the relative time of the clock and observer using time dilation due to the velocity of the clock or by placing the clock at a different gravitational potential. But this is easily avoided by requiring that the relative velocity and gravitational potential of the clock and observer remain zero. Provided we do this there is no mechanism known for the times of the clock and observer to differ.

Option (2) is linked to conservation of energy by Noether's theorem. Without going into the gory details the fact the laws of physics don't change with time implies conservation of energy and vice versa. Since no experiment has ever observed non-conservation of energy most of us believe energy is conserved and therefore that the laws of physics do not change with time. This implies the clock ticks regularly.


There is really no such "thing" as time. It is not matter, not energy. It is only really a concept, used to calibrate the period between one event & another. As I assume you are you are saying, clocks only move or count at a precise rate & indicate time. You can't sample time & measure it like you can, pressure or flow, with a meter. To determine the extreme accuracy of atomic clocks, I assume they ran several of them for a year or more & noted that the slowest & fastest were about 2uSec. apart in a year. The resonant frequency of the caesium atoms in the middle one (9192631770Hz) would have been used for the standard second. Seeing as time was made to fit the clock & they only deviate + - 1uS in a year, they are assumed to be accurate to a second in a million years.


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