8
$\begingroup$

The behavior of an electron (and other tiny things) is said to be probabilistic because we can't say where an election will be when we measure it, but only where it will probably be. As I understand it, the Heisenberg uncertainty principle says the more we know about momentum the less we know about location and vice versa.

Is there some property of nature that makes the behavior of an electron random, or does it simply appear random to us because our ability to predict its location in the future is limited by our inability to determine both momentum and location in the present? Or, as seems likely to me, is it simply impossible for us to know whether the behavior is random because we are limited in our ability to observe the details of what is happening (again because of Heisenberg)?

UPDATE: What I really asking is whether we have an answer to the question, "Does God play dice with the universe."
a. Yes - Even if we knew everything about the state of the universe and could violate HUB by knowing both momentum and position precisely we still wouldn't be able to predict the future. b. No - It's a clockwork but we'll never be able to make predictions because there is a limit to know we can know about the current state of the universe. c. The question is unanswerable because the dice/clockwork are so small that we can't ever see them and the behaviors we are capable of observing are the same regardless.

Improve this question
$\endgroup$
7
  • $\begingroup$ Why do people always capitalize Heisenberg uncertainty principle? Is it standard to do so in some languages? $\endgroup$ – DanielSank Oct 4 '15 at 5:13
  • $\begingroup$ @DanielSank because of the acronym HUP, which is commonly used $\endgroup$ – anna v Oct 4 '15 at 5:36
  • $\begingroup$ @annav It's interesting: I never saw HUP before using this site. $\endgroup$ – DanielSank Oct 4 '15 at 6:37
  • $\begingroup$ @DanielSank it is in the disambiguation of wikipedia en.wikipedia.org/wiki/HUP $\endgroup$ – anna v Oct 4 '15 at 7:40
  • $\begingroup$ I supposed it was seen as a proper name, which are usually capitalised in English ie like Newtons Laws of Motion $\endgroup$ – Mozibur Ullah Oct 4 '15 at 15:30
7
$\begingroup$

The Heisenberg Uncertainty Principle (HUP) holds for special observables, as energy and time, space and momentum, ..

To every observable there corresponds a quantum mechanical operator. Quantum mechanical operators either commute or not commute, and are seen in the commutation relationships. Observables that do not commute are what the HUP is about.

It is the HUP that characterizes the probabilistic behavior of elementary particles and the framework of physics when sizes become small enough that h_bar is significant enough to be seen in the behavior of observables, like momentum and location.

Is there some property of nature that makes the behavior of an electron random,

There are effective random distributions in classical statistical mechanics, and these are defined by the gaussian distribution and the standard deviation that describes the randomness.

or does it simply appear random to us because our ability to predict its location in the future is limited by our inability to determine both momentum and location in the present?

It appears random but the distribution is not a gaussian with its standard deviation giving the error. The distribution is strictly defined by the quantum mechanical equations that give the solutions for the specific boundary conditions.

Or, as seems likely to me, is it simply impossible for us to know whether the behavior is random because we are limited in our ability to observe the details of what is happening (again because of Heisenberg)?

With our measurements we measure the probability distributions and see that they are not gaussian, so we know that there is no randomness. The distributions fit the calculations from the quantum mechanical solutions.

a. Yes - Even if we knew everything about the state of the universe and could violate HUP by knowing both momentum and position precisely we still wouldn't be able to predict the future.

This statement is true for classical statistical mechanics, as h_bar there is effectively 0, because of the immense complexity of the ~10^23 molecules per mole. We would still have to work with the gaussian probabilities.

If there exists a deterministic underlying layer below quantum mechanics, the same would hold true, the complexity would be such that the probabilistic form calculated and validated by quantum mechanics would have to hold . A number of physicists are working on this, not popular, direction as 't Hooft who has also contributed on discussions of his proposals on this site.

The arguments against quantum mechanics being an emergent level from a deterministic one come from space time considerations.

b. No - It's a clockwork but we'll never be able to make predictions because there is a limit to know we can know about the current state of the universe.

Physics never says never for new discoveries.

c. The question is unanswerable because the dice/clockwork are so small that we can't ever see them and the behaviors we are capable of observing are the same regardless.

same as in b.

Improve this answer
$\endgroup$
8
  • $\begingroup$ This is getting the most upvotes (1 so far) though I don't fully understand the part about commutation relationships. What I understand of the second part is that you're saying if the underlying mechanism were deterministic rather than random then we would expect a Gaussian distribution but we don't see a Gaussian distribution therefore the observed behavior is not caused by a deterministic mechanism. I'm not convinced. I can think of a number of ways that a Gaussian distribution can map to another distribution and thus be indistinguishable from that other distribution. $\endgroup$ – Readin Oct 8 '15 at 2:37
  • $\begingroup$ Your question before edit asks about randomness, not about determinism, they are two entirely different things. Statistical mechanics emerges from a deterministic system and it does give gaussians. The contrary is not true. Yes, this answer does not exclude some underlying deterministic process, but it excludes randomness as defined for errors and standard deviations. It has been found that probabilistic quantum mechanics fits the data. Whether the quantum dynamics emerges from underlying determinism is another story. $\endgroup$ – anna v Oct 8 '15 at 3:03
  • $\begingroup$ Perhaps my terms have different meanings in physics. "Random" to me means the number just pops up without any reason for doing so - but does obey certain rules of probability "Deterministic" means that the numbers have a reason for coming up exactly as they to. To me that's equivalent to saying "random" means even if we know all the rules and the exact state of the system we can't predict the future precisely while "deterministic" means if we know all the rules and the exact state we can predict the future precisely. Am I using the words wrong? If so what words should I use? $\endgroup$ – Readin Oct 8 '15 at 4:16
  • $\begingroup$ Yes, you are using the words wrong for physics. Classical statistical mechanics comes from a completely deterministic level, that of individual particles, and still the huge numbers involved introduce randomness which appears in the Gaussian distribution of the probabilities for velocities etc and that is the definition of randomness. When measuring something, a velocity for example, if the accumulation of measurements does not display randomness ( gaussian) we know that there exists a physical reason for that, an energy source in one part of the gas, for example. $\endgroup$ – anna v Oct 8 '15 at 5:48
  • $\begingroup$ Deterministic means that one can write a function that will contain all the variables of the system and describe it. This is impossible for large numbers, as the classical case shows. An underlying deterministic equation for quantum mechanical systems can be written for few particles, but again large numbers will need a probabilistic interpretation, in the case of quantum mechanics the probabilities have to follow the solutions, because they have been verified over and over again, i.e. HUP and everythin. Physics is consistent from one layer to the higher one. $\endgroup$ – anna v Oct 8 '15 at 5:51
0
$\begingroup$

There is something going on. Which is that certain observables are fundamentally incompatible. That means firstly that you can do an experiment for one observable or for another observable, but you can't do an experiment for both observables at the same time.

And what's worse if you did an experiment for A then one for A again and then one for B the two results for A will always agree which means after doing the A experiment the result is clearly in a state that yields a particular result for the A experiment. Remember this, doing the A experiment makes the result clearly in a state that yields a particular result for the A experiment.

However if you instead first do experiment A then do experiment B and then finally do experiment A again its different. The two A experiments can yield different results when the observables A and B are incompatible observables. The order matters. So the first A experiment left it in a state that yields a particular results for the A experiment but doing the B experiment changed it to a different state. So doing these experiments changes the state.

That's what is going on. There are states and observables and sometime doing an experiment for one observable will change the state. And that is unavoidable. And it isn't about some lack of knowledge. The state says everything we know, if we knew more we'd include that to give more or different or better information about what happens in an experiment. The knowledge we get is from the experiments but the experiments change the states and this is unavoidable once the order you do them changes the results.

The behavior of an electron (and other tiny things) is said to be probabilistic because we can't say where an election will be when we measure it, but only where it will probably be.

That's not right. We state the frequencies we will get different results in an experiment if we perform each of the various experiments. Getting certain results under a certain experiment is different than being somewhere (or having a certain momentum) and having that be passively revealed.

As I understand it, the Heisenberg uncertainty principle says the more we know about momentum the less we know about location and vice versa.

Again, no. The principle says that a state that gives a low enough spread of results for position experiments will give a higher spread of results for momentum experiments. The state tells us absolutely everything there is to know about the system, not a lesser or a greater amount, a perfect everything there is to know. It's just that some states give larger or smaller spreads for some experiments. The experiments also change the state. Both follow from the fact that these experiments are required to change the states in order to have the order you do the experiments matter.

The initial state could be the kind that changes into new position states with a wide spread (changes when we do position experiments). Or the initial state could be the kind that changes into new momentum states with a wide spread (changes when we do momentum experiments). But the state itself can't change to a small spread of both any more than a note can have a short duration and a narrow range of pitch.

Is there some property of nature that makes the behavior of an electron random,

It has to do with how you change the state. Its similar to when a single celled organism divides into two symmetrical parts. The two parts can then produce different results even though they have a common past, and the differences accumulate and build up and get larger from how they interact with the rest of the world. Its how you interact and change in response to the experiment that produces the different results. The experiment is specifically designed to have different states interact differently. Thus the original state gets separated into parts each of which interacts differently until the descendants of the original wave eventually think of themselves as the only result. And the singularity of the outcome becomes a matter of perspective, each part having the perspective it is the only part.

or does it simply appear random to us because our ability to predict its location in the future is limited by our inability to determine both momentum and location in the present?

It is wrong to think it had a position or a momentum in the present. If that worked we absolutely would have made a theory like that. It didn't. So we made a theory with states. It had a state and the state evolved. The state evolves differently based on the experiments we do and the experiments change the states. And they do so by separating the state into pieces that act differently and eventually and sometimes become separated from each other and entangled with other things so that their own perspective is that they are the singular outcome of the earlier combined state.

Or, as seems likely to me, is it simply impossible for us to know whether the behavior is random because we are limited in our ability to observe the details of what is happening

We are limited. But not by a lack of knowledge. You are limits to your own perspective based on your own state. You are limited to the results of the experiments you are connected to. And you can't blame your lack of ability to predict the future any more than that single celled organism can be blamed for not predicting whether it went north or south after it divided. If you let it do the dividing experiments many times and give it a method to accumulate the relative frequencies of its experiences then the statistical predictions will be spot on.

And the singularity of the individual experience isn't a failing for a thing that divides and has each new part be able to act like it is a whole.

If you look at the Schrödinger equation description of what happens in a measurement, the state does split, and eventually in some situations has the various parts have a perspective where it is the whole state.

Improve this answer
$\endgroup$
4
  • 2
    $\begingroup$ This doesn't seem to answer the question. Perhaps the answer should directly address the asker's question about whether or not the apparent randomness is due to lack of knowledge, e.g. by mentioning Bell violations. Just stating "Also, it isn't about some lack of knowledge" is, unfortunately, begging the question. The information here is correct it just doesn't really answer the question. $\endgroup$ – DanielSank Oct 4 '15 at 5:21
  • $\begingroup$ @DanielSank I think Bell's inequality is what begs the question with the whole mistaken idea that these experiments are "measurements" that reveal a preexisting property. And I think the question itself begged the real question by assuming you have positions and momentums. You have states that give various results to experiments and the experiments themselves change the states. The lack of commutativity is the entire issue. $\endgroup$ – Timaeus Oct 4 '15 at 5:25
  • 1
    $\begingroup$ If you'd like to discuss the Bell violation issue in the chat room, please do, and ping me there. I'd be interested to hear the details of your point of view. $\endgroup$ – DanielSank Oct 4 '15 at 5:32
  • $\begingroup$ @DanielSank I made a better (and much longer) answer. $\endgroup$ – Timaeus Oct 4 '15 at 5:57
0
$\begingroup$

Some history here might be useful; one of the oldest cosmological theories that we have was developed by Ionian philosophers which was an atomic theory; in that theory uncertainty as in random motion was taken as something fundamental (they called it the clinamen which is usually translated as swerve).

This shows that pure determinism, physically, is something new; and it arose with Newton.

There is a speculative theory of QG, called causal sets which takes uncertainty of motion, which they also call the swerve (presumably in honour of the earlier work) as something basic, and not derivative.

Improve this answer
$\endgroup$
-1
$\begingroup$

First and foremost we have to understand that if we are deriving laws of nature then our primary assumption is that the natural phenomenons are not random. If they are random, it would be impossible to say anything about them.

Now let's come back to Quantum Mechanics. There is nothing random about the motion of electrons or any other subatomic particles until you are trying to observe it. Their behavior is completely determined by Schrodinger equation. Now Schrodinger equation gives the time evolution of a wave associated with the system. Born interpreted this wave as the distribution of probability amplitude and according to this theory if you take a large ensemble of identical systems then the results of some particular experiment may differ from system to system, but their relative abundance will be completely governed by Schrodinger equation.

So yes, if you take one system and ask "where would I find this particular particle?" quantum mechanics cannot give you perfect answer but can only say "there is a large probability that you will find it at x=something". Your experiment may or may not give this answer, but that does not mean it is random. Because had you done the experiment repeatedly (but each experiment must be done after sufficient time gap from the former, better still they should be performed in a different identical system), you will find that quantum mechanics will give you almost exact data about the relative probability of finding the particle at some state or the other.

Improve this answer
$\endgroup$
3
  • $\begingroup$ If I repeat an experiment preparing and measuring a quantum system, I find that the measured results are random and (to a large degree, up to systematic errors) uncorrelated. We do indeed have a theory describing this which includes a deterministic part (Schrodinger's equation or whatever version you like), but there is also a part involving probability and so far that part remains an absolutely essential element in the theory. Emphasizing that the probability distribution is deterministically predicted by the theory seems like an odd argument that the "phenomenons [sic] are not random". $\endgroup$ – DanielSank Oct 4 '15 at 5:28
  • $\begingroup$ @DanielSank It is almost a disservice to say that there is a part with probability. For instance, every time you switch between incompatible observables (which is all the time) you are supposed to make a new sample space for your probability if you want the results to be random variables (which commute). And also we have zero evidence that the deterministic evolution is ever wrong. But anyway I think the answer above just meant they follow a distribution as opposed to truly random. Truly random, e.g. like the motion of a particle at rest at the top of Norton's dome that isn't perturbed. $\endgroup$ – Timaeus Oct 4 '15 at 6:08
  • 1
    $\begingroup$ Respoonding to one thing at a time: "And also we have zero evidence that the deterministic evolution is ever wrong". It's "wrong" precisely when you make a measurement. Decoherence gets us a diagonal density matrix but then we have to explain why I actually see one specific outcome, and we can't do that without probability. That's why I say the theory has two parts: the unitary evolution (which as you say seems to be pretty darn accurate!) and the "collapse" part (or whatever you want to call it). It's not a disservice to tell the truth about the data being probabilistic. $\endgroup$ – DanielSank Oct 4 '15 at 6:39

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