Not to defend Phrak, but i have a related question

[rbj]

clocks measure time by counting events that we believe are periodic with whatever this quantity we call "time" is.

so, how exactly do we decide that, whatever "time" is, there is an equal amount of it between the onsets of "periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the cesium 133 atom"?

is that an assumption regarding the stability of such radiation? or are there other natural periodic phenomena that we compare this radiation to, and discover that, to a very high degree of precision, they all have a number of oscillations that are extremely close (so close that we cannot measure or sense an error) to precisely proportional to each other (given the same and simultaneous period of counting), that we believe them all to be oscillating at their individual constant rates w.r.t. time?

i think that this is essentially what any question regarding "what do clocks measure?" should mean. and if the thread was unlocked, i would have added this to that thread rather than start another.

 

[russ_watters]

I don't know the theories that govern how a cesium atom behaves or exactly how a cesium clock works, but with all other clocks that I can think of, the periodic nature of the events are governed by physical theories. A rotating planet and a swinging pendulum have relatively simple physical rules that govern their operation.

For a simple pendulum or spring-mass system (or basically any physical oscillation): http://hyperphysics.phy-astr.gsu.edu/hbase/shm.html
For a rotating planet, there are no equations. The period is the period.

 

[russ_watters]

Here's how a cesium clock works (I really should already know this...): http://tycho.usno.navy.mil/cesium.html

A "cesium(-beam) atomic clock" (or "cesium-beam frequency standard") is a device that uses as a reference the exact frequency of the microwave spectral line emitted by atoms of the metallic element cesium, in particular its isotope of atomic weight 133 ("Cs-133"). The integral of frequency is time, so this frequency, 9,192,631,770 hertz (Hz = cycles/second), provides the fundamental unit of time, which may thus be measured by cesium clocks.

So it's actually just measuring the frequency of light emitted by an excited cesium atom. That the frequency is constant is fundamental to QM.

 

[jtbell]

is that an assumption regarding the stability of such radiation? or are there other natural periodic phenomena that we compare this radiation to, and discover that, to a very high degree of precision, they all have a number of oscillations that are extremely close (so close that we cannot measure or sense an error) to precisely proportional to each other (given the same and simultaneous period of counting), that we believe them all to be oscillating at their individual constant rates w.r.t. time?

I think that's the proper way to consider it. We don't develop new kinds of clocks in isolation and simply assume that they run uniformly. We always compare them with clocks that we already have, either directly, or indirectly by using them to measure the time evolution of the same physical phenomena. If we compare two kinds of clocks and observe a discrepancy, there's obviously a question as to which one is "more correct." If we compare several different kinds of clocks and one of them is "out of line" with the rest, beyond the expected random experimental error, then we should be suspicious of it.

Ultimately I think we should expect that "good" clocks should be the ones that give us the "simplest" results for measurements of the time evolution of physical processes. For example, according to Newton's First Law, an object with zero net force acting on it should travel in a straight line in equal distances in equal time intervals, and not have a measured velocity that oscillates or fluctuates randomly.

 

[rbj]

For a rotating planet, there are no equations. The period is the period.

Russ, I'm a little astonished to read that.

of course a rotating planet is not an absolute standard for time. there are all sorts of things that are known that make it drift slower and wobbly.

now, my question (which I'm not sure if it's the same as Phrak's question or not) is:

just as we have once used astronomical measurements for our time base, we have since concluded that they are not solidly based solely on time (and linear with time). now we use some atomic behavior (emission of radiation) and use that as a time base. now, how do we know that this measures only time and is not counting the periodic behavior of something that is not also drifting or wobbling?

 

[kdv]

clocks measure time by counting events that we believe are periodic with whatever this quantity we call "time" is.

so, how exactly do we decide that, whatever "time" is, there is an equal amount of it between the onsets of "periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the cesium 133 atom"?

is that an assumption regarding the stability of such radiation? or are there other natural periodic phenomena that we compare this radiation to, and discover that, to a very high degree of precision, they all have a number of oscillations that are extremely close (so close that we cannot measure or sense an error) to precisely proportional to each other (given the same and simultaneous period of counting), that we believe them all to be oscillating at their individual constant rates w.r.t. time?

i think that this is essentially what any question regarding "what do clocks measure?" should mean. and if the thread was unlocked, i would have added this to that thread rather than start another.

I will be probably be shut off since I have apparently been deemed a crackpot (there is a difference between turning down people who prentedn having new theories without having understood the present accepted theories, and someone who wants to have an open discussion about new ways of thinking about things, but apparently it's hard for some people to be able to tell the two apart. For the record, I was deemed a crackpot because I observed that any time measurement essentially amounts to a measurement of the position of something because in physics, all measurements are position measurements. If anyone is not close minded and wants to offer counterexamples and give me the chance to reply, I will be glad to do so by private e-mail since some administrators refuse to allow a constructive and respectful discussion about this).


Anyway, you are right that there is no absolute measure of time. In the end, we never measure time directly, we always compare two physical processes to one another. For example we will compare the instants when a pendulum is back at a given position to the number of photons produced by the nuclear decay of some unstable nucleus. So we are essentially always comparing standards of time against one another.
And before prevect or Russ Waters lock off the thread because they don't think this is scientific enough for them, let me point out that this is an essential point of view adopted in loop quantum gravity (so if I am a crackpot, I am in the good company of Lee Smolin, Carlo Rovelli and many other well-known researchers).


Actually, if you read the book on quantum gravity by Rovelli, this is one of the key points made. He even discuss what he calls the "timeless pendulum" example. Where you consider two pendulum (pendula I guess!). Normally, one would say that th emotion of each pendulum depends on a mysterious time variable. But this time is never observed in itself! The only thing that we actually observe is the motion of the two pendula. So one way to describe the evolution of the pendula is to actually eliminate completely the time variable and to give the position of one pendulum as a function of the position of the second pendulum. And this is a key idea in loop quantum gravity: that we may have to eliminate completely the concept of time before being able to quantize properly gravity.

In actuality, one never measures directly "time". Time is actually a convenient mathematical concept that is introduced in order to simplify both the maths and the concepts of physics. In the end, one always compares two different physical phenomena (for example a pendulum oscillating and the regular dropping of water dropplets or the decay of some unstable nucleus to the oscillation of a quartz crystal and so on).

Going back to your question, the only way to know if there is stability in some peridoic phenomenon is to compare to several other physical phenomenon and see if all of them are consistent over many many cycles.

 

[russ_watters]

Russ, I'm a little astonished to read that.

of course a rotating planet is not an absolute standard for time. there are all sorts of things that are known that make it drift slower and wobbly.

All clocks have errors. If you don't like the Earth, use a hypothetical planet without a moon to slow it down. Or better yet, a pulsar. Pulsars are highly stable. This objection is a non sequitur.

just as we have once used astronomical measurements for our time base, we have since concluded that they are not solidly based solely on time (and linear with time). now we use some atomic behavior (emission of radiation) and use that as a time base. now, how do we know that this measures only time and is not counting the periodic behavior of something that is not also drifting or wobbling?

The Earth's rotation is not a stable clock because the Earth's rotation is slowing and it is slowing because it has something pulling on it. Take away the moon and it would keep much better time. This is, in any case, irrelevant to the theoretical discussion. A clock based on the principle of an object rotating at a stable rate would keep good time. And such objects exist that rotate at a more stable rate than the Earth. The Earth was used as an example simply because it has been used as a clock in the past. (and did a fine job at that for most of human history)

Clocks based on simple harmonic motion (such as pendulum clocks) and clocks based on QM have the laws of physics to control their oscillations/rates. The only way their clock rates can drift beyond a certain associated random error is if the laws of physics are wrong. And like jtbell said, when a new clock is invented, it is tested against other clocks to gage its accuracy relative to them.

I'm really not sure of the point of this line of questioning. Can you tell me why a clock with true periodic behavior would drift beyond experimental error? Isn't that a contradiction in terms?

 

[russ_watters]

The argument you guys keep bringing up about the Earth's slowing rotation rate is a non sequitur, but to avoid the issue entirely, from now on we should talk only about pulsars as an example of an object rotating at a stable enough rate to be used as a clock.

As cosmological clocks they can rival in their long-term stability cesium clocks on earth.

http://www.ras.ucalgary.ca/SKA/science/node22.html

 

[russ_watters]

For the record, I was deemed a crackpot because I observed that any time measurement essentially amounts to a measurement of the position of something because in physics, all measurements are position measurements.

At face value that idea seems pretty clearly wrong and in any case, it isn't appropriate for this thread, but I'll bring the issue up with the other moderators to see what we can do with it.

Anyway, you are right that there is no absolute measure of time. In the end, we never measure time directly, we always compare two physical processes to one another. For example we will compare the instants when a pendulum is back at a given position to the number of photons produced by the nuclear decay of some unstable nucleus. So we are essentially always comparing standards of time against one another.

So what? You guys keep saying these things as if they are somehow unique to time and/or somehow make time not a real thing or a more limited concept than position/length. How are these measurement limitations any different from our limitations in measuring length? They most certainly do not imply any problems with the concept of time and how it is used in physics. In fact, a cesium clock is better as a standard of length than a platinum bar sitting in a vacuum chamber somewhere. That's why our length standard is based on our time standard and not the other way around. So what is your point?

 

[Schrodinger's Dog]

Here's how a cesium clock works (I really should already know this...): http://tycho.usno.navy.mil/cesium.html So it's actually just measuring the frequency of light emitted by an excited cesium atom. That the frequency is constant is fundamental to QM.

I don't see why, I rather naively assumed it was governed by half lives, until I realized this might be a tad difficult to do in practice and actually looked it up, at which point I had to completely edit my post in Frak's infamous time thread to include a short googled excerpt.

Essentially it raises the point how accurate is the most accurate form of measurement? Well since ultimately it relies on the motion of the sun and the planets and in turn on hours and minutes and in turn on mili and nanoseconds. Only as accurate as it's relationship with that. The margin of error for the caesium atom is so tiny that it loses a second every few million years. To all intents and purposes that is as accurate as it needs to be.

 

[my_wan]

Essentially it raises the point how accurate is the most accurate form of measurement?

In essence we already know the answer to this question. It's defined by the Heisenberg Uncertainty relation. There are two notions of time being discussed here both of which is contained in the notion of accuracy.

The first and more reasonable is how accurate the measurement of time can be even in principle. Suppose we did assume that at very tiny increments in time that the increments we were counting fluctuated wildly wrt the background of events. Yet when averaged over the relative eons of a second these fluctuations are totally meaningless. Any inaccuracy associated with these hypothetical fluctuations would not accumulate over large spans of time. Accuracy is therefore well maintained. Accuracy of even these tiny increments of time can be maintained by averaging over locally simultaneous measurements of events.

The second and sillier is absolute time that supposedly exist independent of the accuracy of our clocks. If the point is that time doesn't exist independent of events capable of measuring it, well duh. Suppose I had a magic button and pressing this button stopped time throughout the Universe for 1 hour. When time started again you could not say that our clocks are not accurate because it failed to notice that hour. Wrt the laws of physics the event was totally imaginary, even in principle. Even the definition of stopped for 1 hour is imaginary. This even has a more severe problem when you ask when the button was pressed wrt different observers. There is simply no way to define an event like this in a globally consistent manner. Demanding this kind of silliness gets threads locked after explanations are refused. It is just as imaginary to speak of the accuracy of a clock wrt such a thought experiment.

 

[Schrodinger's Dog]

I agree, the only way time has meaning to us is in context with our frame of reference. Basically, any other method of determining things which isn't observable or measurable by us, is useless because it is unobservable.

 

[rbj]

just want to say that i disagree with Russ that the issue of wobbly astronomical methods is non sequitur. the issue is just as we first assumed (a couple hundred years ago) the particular astronomical method was solid and have since found that counting the Earth's rotations was not precisely the same as measuring equal periods of "time" (whatever that is), do we not continue to have the same philosophical issue with the new and current methods, whether it be a pulsar or a cesium atom?

i think jtbell picked up on my question. it gets down to, what exactly is time, and what exactly do clocks measure since this quantity we call "time" is a bit elusive to get a perfect grip on (even in one's own unaccelerated frame-of-reference out in a near vacuum and reasonably flat space-time).

from here on, i keep my mouth shut about this.

 

[my_wan]

just want to say that i disagree with Russ that the issue of wobbly astronomical methods is non sequitur.

I will conditionally agree here. In principle if perturbations of planetary orbits was precisely known then after such corrections these orbits would be as good a clock as any. This seemed to basically be russ's point in mentioning pulsars. However, the OP was in reference to absolute accuracy not utility.

The only way we can talk about accuracy of clocks in any absolute sense is when measured wrt the laws of physics. To this end we have an invariant called the space-time interval. This means that wrt physical laws we, at least in principle have an absolutely perfect clock. A light clock. By it's very definition wrt physical laws any disagreements between clocks is not an inaccuracy but a difference in actual rates of time. Of course we still have the Uncertainty principle, transient local gravitational effects, thermal effects on equipment, limitations on event rates to count, and less than perfect methods of detecting and counting events. It is wrt to an idealized light clock (i.e., wrt physical laws) that the accuracy of clocks is judged.

http://www.sciencedaily.com/releases/2008/02/080214144459.htm

 

[Marcwhydothe]

rember numbers are infinite measurements are infinite its the impact on the final outcome that really matters.
For any mathematical model for a universal law it must predict the future events with high amount of accuracy then it is debated why it makes sense to us then finaly it is agreed upon.
time can be measured to infinite and back but it doesn't really matter because in the real world when we use time in our caculations it usaly predicts the event that occur in our for example solar system.
Time may not exist but its a great way to constrain the universe to determine future events so that we can understand them through correct predictions! lol
( this is just random thought I had lol)
The really weird thing is without time wed be lost its its kinda like gravity it puts everything into motion. if no time had passed then nothing would happen.
By the way I am just ranting answer do really know if this sounds good lol
remeber time is measured by the object in are universe right we are object in the universe constrained by time and therefore that is why time is useful to us and works. But if we weren't constrained by timehmmm... ( It would be really weird)

 

[Schrodinger's Dog]

rember numbers are infinite measurements are infinite its the impact on the final outcome that really matters.
For any mathematical model for a universal law it must predict the future events with high amount of accuracy then it is debated why it makes sense to us then finaly it is agreed upon.
time can be measured to infinite and back but it doesn't really matter because in the real world when we use time in our caculations it usaly predicts the event that occur in our for example solar system.
Time may not exist but its a great way to constrain the universe to determine future events so that we can understand them through correct predictions! lol
( this is just random thought I had lol)

True but it's already been said, although that's a nice way of putting it.

If there is no objectivity in science then science is flawed.

 

[russ_watters]

The second and sillier is absolute time that supposedly exist independent of the accuracy of our clocks. If the point is that time doesn't exist independent of events capable of measuring it, well duh. Suppose I had a magic button and pressing this button stopped time throughout the Universe for 1 hour. When time started again you could not say that our clocks are not accurate because it failed to notice that hour. Wrt the laws of physics the event was totally imaginary, even in principle. Even the definition of stopped for 1 hour is imaginary. This even has a more severe problem when you ask when the button was pressed wrt different observers. There is simply no way to define an event like this in a globally consistent manner. Demanding this kind of silliness gets threads locked after explanations are refused. It is just as imaginary to speak of the accuracy of a clock wrt such a thought experiment.

Yes, that's the sort of silliness I see in these threads.

 

[russ_watters]

just want to say that i disagree with Russ that the issue of wobbly astronomical methods is non sequitur. the issue is just as we first assumed (a couple hundred years ago) the particular astronomical method was solid and have since found that counting the Earth's rotations was not precisely the same as measuring equal periods of "time" (whatever that is), do we not continue to have the same philosophical issue with the new and current methods, whether it be a pulsar or a cesium atom?

The point is that there are no philosophical issues here. We merely found that the clock we were using had intrinsic error in it. It's the same as if you found the watch you just bought didn't keep good time. You'd throw it out and buy a new one. That doesn't suggest anything at all about the nature of time itself.

i think jtbell picked up on my question. it gets down to, what exactly is time, and what exactly do clocks measure since this quantity we call "time" is a bit elusive to get a perfect grip on (even in one's own unaccelerated frame-of-reference out in a near vacuum and reasonably flat space-time).

Why people have such trouble with time is beyond me. The concept is not elusive, it seems that people simply refuse to accept the physics behind it. The concept is quite well understood by scientists.

 

[kdv]

The point is that there are no philosophical issues here. We merely found that the clock we were using had intrinsic error in it. It's the same as if you found the watch you just bought didn't keep good time. You'd throw it out and buy a new one. That doesn't suggest anything at all about the nature of time itself. Why people have such trouble with time is beyond me. The concept is not elusive, it seems that people simply refuse to accept the physics behind it. The concept is quite well understood by scientists.

Do you know anything about loop quantum gravity? If the concept of time si completely obviosu to you, than I guess far superior to any of those reserachers trying to build a consistent hamiltonian formulation of quantum gravity.

But calm down. If it's not possible to have a respectful, open-minded and civilized exchange of ideas here, we will stop talking about it and find some other forum to discuss these ideas.

we got the message.

 

[russ_watters]

rember numbers are infinite measurements are infinite its the impact on the final outcome that really matters.

Time may not exist but its a great way to constrain the universe to determine future events so that we can understand them through correct predictions! lol

remeber time is measured by the object in are universe right we are object in the universe constrained by time and therefore that is why time is useful to us and works.

Yes. Science can provide us nothing more or less than a model of reality that either works or doesn't. If it doesn't work, we endevour to find one that does. On the issue of time, physics provides us with an extremely accurate measurement stick via time. Time is, in fact, more useful and accurate as a measuring stick than length itself - which is why length is now defined in terms of time.

 

[russ_watters]

Do you know anything about loop quantum gravity?

No, I don't.

If the concept of time si completely obviosu to you, than I guess far superior to any of those reserachers trying to build a consistent hamiltonian formulation of quantum gravity.

So far, the questions being asked and issues being proposed about time in these threads are quite rudimentary. I have yet to see a real issue here with how time works for us at present. At least, no one has actually brought up any real issues with our understanding of time except for backhanded mentions of LQG without explanation/citation of real problems it would address. Again, I ask: what is your point? Does LQG explain anything about the nature of time that would suddenly make (for example) our GPS system more accurate?

But calm down. If it's not possible to have a respectful, open-minded and civilized exchange of ideas here, we will stop talking about it and find some other forum to discuss these ideas.

I'm quite calm. I do (and this forum does), however, have a relatively short tolerance for idle speculation. And I get frustrated and a little irritated when people make backhanded insinuations without explanation of what they are actually getting at. It implies deception to me.