Differing first-principle models for Maxwell-Boltzmann statistics?

[greswd]

Let's consider a simple scenario in Maxwell-Boltzmann statistics: 6 identical but distinguishable particles, and 9 quanta of energy, 9 indivisible units, to be distributed among the particles.The first model is like that of the wheels on a combination lock, or should I say "permutation lock".

There are 6 wheels, one for each particle, and the numbers run from 0 to 9, since there are 9 units of energy available.

The number on each wheel represents the amount of energy each particle possesses. Since there are 6 wheels, there are a total of a million permutations.

However, there are only 9 units of energy in total, so if the 6 numbers do not add up to 9 in total, it is an invalid permutation.

Considering only the valid permutations, the odds of a particle being in the ground state, having zero units of energy, is the highest of all, and it only decreases subsequently for each incremental energy level.
Now for the second model, 9 dice.

One die for each unit of energy, and of course 6 sides, one for each particle.

You can imagine rolling all 9 dice, with the number on each die indicating which particle the unit of energy the die represents belongs to.

This time, there's no need to remove invalid permutations, all are valid.

This model produces quite different results from the first one, as the odds of a particle having one unit of energy is the highest, and the odds of having two units of energy is even higher than the odds of having no energy at all.Maxwell-Boltzmann statistics follows the first model, and I'm curious as to why nature is of the first model rather than of the second.
Because intuitively, it seems to me like the second model would be more likely to occur. It's simpler in a way, no assumptions need to be made about invalid permutations.

 

[BvU]

You can imagine rolling all 9 dice, with the number on each die indicating which particle the unit of energy the die represents belongs to.

So state 1 1 1 1 1 1 1 1 2 occurs 9 times (any of the other dice can be 2 instead of 1) ?

 

[greswd]

So state 1 1 1 1 1 1 1 1 2 occurs 9 times (any of the other dice can be 2 instead of 1) ?

yes, so maybe that state should have a statistical weight of 9?

 

[BvU]

What would it get in the permutation lock scenario ?

 

[greswd]

What would it get in the permutation lock scenario ?

It is, as you've mentioned, a state, so I think it's weight is 1, out of 2002.

While for the dice it is 9 out of 10077696.

 

[Dale]

You can imagine rolling all 9 dice, with the number on each die indicating which particle the unit of energy the die represents belongs to.

This time, there's no need to remove invalid permutations, all are valid.

This is not exactly correct. Although all permutations are valid, they are not all distinguishable. For example {1,1,1,1,1,1,1,2,2} refers to the same state as {2,2,1,1,1,1,1,1,1} and {2,1,1,1,1,1,1,1,2}. The particles are distinguishable, but the energy levels are not. If you go through and eliminate all of the invalid permutations of locks and all of the indistinguishable permutations of dice then you are left with 2002 sets in both cases.

The equipartition theorem basically states that all of these 2002 distinguishable/valid sets are equally probable. You could make an arbitrary number of models describing how many indistinguishable sets correspond to each distinguishable set, but that doesn't affect the physics. So physically both scenarios are the same since they both wind up with the same number of sets. In one case you have to remove invalid cases and in the other you have to remove indistinguishable cases.

It is, as you've mentioned, a state, so I think it's weight is 1, out of 2002.

While for the dice it is 9 out of 10077696.

In both cases it is 1 out of 2002.

 

[BvU]

Conversely, all 9 energy bits for the first particle gets one out of 10077696 with the dice and 1 out of 2002 with the lock

The energy bits are indistinguishable, so you'd have to divide each of the 10077696 combinations by the number of ways such a combination can occur to get its weight.

(ah, Dale to the rescue while I waste my time trying to get an elegant expression for this dice combination weight )

 

[greswd]

The equipartition theorem basically states that all of these 2002 distinguishable/valid sets are equally probable. You could make an arbitrary number of models describing how many indistinguishable sets correspond to each distinguishable set, but that doesn't affect the physics.

thanks, so, what's the root, fundamental basis or effect that leads to equipartition probabilities and nature not following the dice distribution?

Is it like some fundamental rule/effect which describes distinguishability affecting probabilities?

Because it seems like the combination lock method isn't a direct, sole logical given, and that there is some fundamental physical principle or effect which makes it so.

 

[BvU]

You can't expect nine dice to mimic nine units of energy ...

 

[greswd]

You can't expect nine dice to mimic nine units of energy ...

But six wheels of a lock demonstrate the mathematics of the phenomenon very well.I think a more direct, clear and physical visual would be a beer-pong set-up.

Imagine 6 cups representing the 6 particles, and 9 ping-pong balls representing the units of energy, which are hurled at the cups.

Assuming the balls are ghostly and pass through each other instead of bouncing off each other, it will produce the same results as that of the dice.

 

[Dale]

thanks, so, what's the root, fundamental basis or effect that leads to equipartition probabilities and nature not following the dice distribution?

I think that it is called the identity of indiscernibles. Two states that are completely indistinguishable are the same state. I believe that it is simply an assumption that seems to work. Apparently it has a very powerful usage in quantum mechanics where it is the cornerstone of the statistical distributions used to represent particles.

The reason that the combination lock works and the dice does not is that you specified that the particles were distinguishable and the energy levels were not. The combination lock respects that distinction and the dice do not. If you had indistinguishable particles then the combination lock metaphor wouldn't work either.

 

[greswd]

I think that it is called the identity of indiscernibles. Two states that are completely indistinguishable are the same state.

I believe that it is simply an assumption that seems to work.

wow, that's pretty interesting. I think most would imagine a "beer-pong distribution" as the simplest method of probabilistically distributing energy among particles. but nature appears to have other ideas.

 

[Dale]

I think most would imagine a "beer-pong distribution" as the simplest method of probabilistically distributing energy among particles. but nature appears to have other ideas.

As long as the balls are indistinguishable and the cups are distinguishable it will work fine.

 

[greswd]

As long as the balls are indistinguishable and the cups are distinguishable it will work fine.

Considering actual, macroscopic ping-pong balls, I'm not sure how indistinguishability would apply.

Whether all 9 balls are white, or each painted with a different color, it would have no effect on the overall distribution. A layer of paint can't change anything.

 

[greswd]

The energy bits are indistinguishable, so you'd have to divide each of the 10077696 combinations by the number of ways such a combination can occur to get its weight.

As long as the balls are indistinguishable and the cups are distinguishable it will work fine.

Considering actual, macroscopic ping-pong balls, I'm not sure how indistinguishability would apply.

Whether all 9 balls are white, or each painted with a different color, it would have no effect on the overall distribution. A layer of paint can't change anything.

Or, a reverse scenario, if each ping-pong ball is originally of a unique color, we could make them "indistinguishable" by painting all of them white.

And that layer of white paint also can't change anything, the kinematics will remain the same.

 

[BvU]

Does someone have a decent expression for the weight of a 9 dice throw ?

 

[greswd]

Does someone have a decent expression for the weight of a 9 dice throw ?

well, there are 10077696 possible permutations in total.

 

[BvU]

I mean: What's the weight if I throw 1 2 3 4 5 6 6 6 6

 

[greswd]

I mean: What's the weight if I throw 1 2 3 4 5 6 6 6 6

it should be 15120

 

[Dale]

Considering actual, macroscopic ping-pong balls, I'm not sure how indistinguishability would apply.

It doesn’t. But IF it did then the beer pong distribution would work. The observed fact that it doesn’t work is due to the violation of the indistinguishability assumption. Clearly, a layer of paint is not what makes a physical system indistinguishable.

 

[BvU]

it should be 15120

So do you have an expression for that ? The number isn't interesting.

 

[greswd]

It doesn’t. But IF it did then the beer pong distribution would work. The observed fact that it doesn’t work is due to the violation of the indistinguishability assumption. Clearly, a layer of paint is not what makes a physical system indistinguishable.

what do you think makes a physical system indistinguishable?

each ball has tiny imperfections and flaws that distinguish it from the others.

however, even if they are somehow 100% perfect copies of each other, it wouldn't affect the kinematics and wouldn't produce different results.
Their trajectories through the air would not change much.

So energy quanta probably have some other properties in addition to their indistinguishability which make them behave differently and follow the lock model.

 

[Dale]

So energy quanta probably have some other properties

I think it is the opposite. They have fewer properties, not additional ones. That is what makes them indistinguishable. Macroscopic systems have too many properties for the identity of indiscernible to play a role.

 

[greswd]

So do you have an expression for that ? The number isn't interesting.

$\operatorname{nCr}\left(9,5\right)\cdot5!$

 

[greswd]

I think it is the opposite. They have fewer properties, not additional ones. That is what makes them indistinguishable. Macroscopic systems have too many properties for the identity of indiscernible to play a role.

The concept of energy quanta is somewhat abstract, but if we replaced all of the ping-pong balls with classical electrons, which behave like tiny ping-pong balls somewhat, I think the result probably wouldn't change.

I think even real electrons in a beer-electron-pong set-up would result in the dice model.

 

[Dale]

I think even real electrons in a beer-electron-pong set-up would result in the dice model.

If that were true then we would be able to observe substantial departures from equipartition

 

[greswd]

If that were true then we would be able to observe substantial departures from equipartition

If we fire 9 electrons, one at a time, at a set-up of 6 holes where the electron has a 1/6th chance of entering anyone hole, we can see from there that it is no different from beer-pong.

 

[Dale]

If we fire 9 electrons, one at a time, at a set-up of 6 holes where the electron has a 1/6th chance of entering anyone hole, we can see from there that it is no different from beer-pong.

If that were true then there would be a violation of equipartition. I don't have any evidence to suggest that is correct, do you?

 

[greswd]

If that were true then there would be a violation of equipartition. I don't have any evidence to suggest that is correct, do you?

But you can imagine a fair set-up, with a 1/6th chance each, like a fair die.
And the electrons are fired one at a time. It is virtually identical to beer pong.

 

[Dale]

It is virtually identical to beer pong.

With the very big difference of distinguishability, right? You are still envisioning that the electrons are indistinguishable, but ping pong balls are not. That has testable consequences.

 

[greswd]

With the very big difference of distinguishability, right? You are still envisioning that the electrons are indistinguishable, but ping pong balls are not. That has testable consequences.

But can you imagine how it would differ?
We set it up with a fair 1/6th chance each, or very closely to perfect fairness.
But somehow the electrons behave differently and don't behave according to those odds.
That would be very strange. Electrons can indeed behave very strangely, or non-classically, but not in this manner.

 

[DrClaude]

If I understand correctly, in the dice / beer pong analogy, you want to label the units of energy, distinguishing which is where?

If that is the case, you need to understand that keep track of the energy is only accounting (see the Feynman lectures). The state of the economy does not depend on which dollar is in which bank account.

 

[greswd]

If I understand correctly, in the dice / beer pong analogy, you want to label the units of energy, distinguishing which is where?

If that is the case, you need to understand that keep track of the energy is only accounting (see the Feynman lectures). The state of the economy does not depend on which dollar is in which bank account.

I mentioned removing all forms of labels in #15, and also mentioned throwing all 9 balls at the same time, albeit "ghostly" balls, in #10.

 

[DrClaude]

I mentioned removing all forms of labels in #15, and also mentioned throwing all 9 balls at the same time, albeit "ghostly" balls, in #10.

Then I must admit I don't understand the discussion.

 

[Dale]

Electrons can indeed behave very strangely, or non-classically, but not in this manner.

I think they do behave non-classically in exactly this manner. At least I am not aware of any evidence of a violation of the equipartition theorem. Are you? You seem very convinced by this, but the consequences would be easily observable.

 

[greswd]

I think they do behave non-classically in exactly this manner. At least I am not aware of any evidence of a violation of the equipartition theorem. Are you? You seem very convinced by this, but the consequences would be easily observable.

But electrons, no matter how non-classical, can't change the 1/6th odds of the fair setup. And its one at a time so two electrons can't interfere with one another.

So it could be that Maxwell-Boltzmann statistics are right but with the wrong explanation.

Or because they don't refer to electrons, but quanta of energy, which somehow behave differently.

 

[Dale]

And its one at a time so two electrons can't interfere with one another.

If you detect after each electron where it went then that makes it distinguishable. If you fire them in one at a time so there is no interference but don't observe their locations until the end then I think they are indistinguishable and would have the corresponding statistics, not the beer pong statistics.

I don't think it works the way you seem to think, and I am not aware of any evidence suggesting it works the way you suggest which I think would be big-news kind of evidence. And you don't seem to be aware of any such evidence either.

It is certainly also possible that I am misunderstanding or misapplying the equipartition theorem, but as far as I know distinguishability has some pretty dramatic physical consequences.

 

[greswd]

If you detect after each electron where it went then that makes it distinguishable. If you fire them in one at a time so there is no interference but don't observe their locations until the end then I think they are indistinguishable and would have the corresponding statistics, not the beer pong statistics.

But if its like the double-slit experiment, with the beer cups being like the slits, the electrons just pass right through the slits, and there is nothing to observe at the end.

Kinda like having cups with the bottoms cut off, the ping-pong balls just fall right through, and at the end of the experiment, all we're left with is 6 empty cups.

 

[Dale]

But if its like the double-slit experiment, with the beer cups being like the slits, the electrons just pass right through the slits, and there is nothing to observe at the end.

I didn't say anything about the double slit experiment. It is not the only experiment that invalidates the notion of counterfactual definiteness. There is no counterfactual definiteness in QM, regardless of if you are describing electrons in wells or photons through slits.

 

[greswd]

I didn't say anything about the double slit experiment. It is not the only experiment that invalidates the notion of counterfactual definiteness. There is no counterfactual definiteness in QM, regardless of if you are describing electrons in wells or photons through slits.

If you detect after each electron where it went then that makes it distinguishable. If you fire them in one at a time so there is no interference but don't observe their locations until the end then I think they are indistinguishable and would have the corresponding statistics, not the beer pong statistics.

noted, not double-slit. what kind of experimental set-up do you think would yield the corresponding statistics?
because I'm wondering what it means to not observe during the firing and observing at the end of it all.

 

[Dale]

noted, not double-slit. what kind of experimental set-up do you think would yield the corresponding statistics?
because I'm wondering what it means to not observe during the firing and observing at the end of it all.

I would think of some sort of potential well created by a circuit where you could turn the individual wells on or off as needed. They would need to be deep enough potential wells that 9 electrons plus or minus would not alter the probabilities. Then you can turn off the wells one at a time and count how many electrons leave each well.

 

[greswd]

I would think of some sort of potential well created by a circuit where you could turn the individual wells on or off as needed. They would need to be deep enough potential wells that 9 electrons plus or minus would not alter the probabilities. Then you can turn off the wells one at a time and count how many electrons leave each well.

That means that somehow the electrons are defying the setup that every well has a 1/6th chance. Interesting, someone should totally conduct this experiment.

 

[Dale]

That means that somehow the electrons are defying the setup that every well has a 1/6th chance.

Actually, no it doesn't mean that. If you go to all of the 2002 distinct configurations you still find that each well is equally likely to contain any given amount of energy. The 1/6 chance is not violated. All that is changed is the distinction if a well with one unit of energy got the first unit or the last unit. In either case it still had the same 1/6 chance to get it but those two scenarios are actually the same scenario.

 

[greswd]

wow, that's pretty interesting. I think most would imagine a "beer-pong distribution" as the simplest method of probabilistically distributing energy among particles. but nature appears to have other ideas.

As long as the balls are indistinguishable and the cups are distinguishable it will work fine.

I've been thinking about this, do you think a system of particles in nature uses a similar model to the "beer-pong scenario" to distribute energies?

it could be a different form of distribution

 

[greswd]

Actually, no it doesn't mean that. If you go to all of the 2002 distinct configurations you still find that each well is equally likely to contain any given amount of energy. The 1/6 chance is not violated. All that is changed is the distinction if a well with one unit of energy got the first unit or the last unit. In either case it still had the same 1/6 chance to get it but those two scenarios are actually the same scenario.

but physically, the electrons land in the wells no differently than the balls in the cups. so I can't imagine how we would be able to arrive at different numerical results