[watsisname]

Fascinating.  I wonder if that might apply to tachyons as well (assuming they exist... which is pretty doubtful).

Also found out my university offers a course in QFT this fall.  Think I might sign up for it. 

[General Battuta]

And Astronomiya, could you describe why reversing time would involve flipping matter and antimatter states as you mentioned earlier?

In QFT, if you imagine a normal particle running backwards in time, it turns out to be exactly mathematically equivalent to an antiparticle running forwards in time, and vice versa.  So, if time suddenly reversed, matter could be said to become what is now antimatter, and vice versa.  It is an interesting interpretation of the underlying physics, and I feel it could very well be reality.  I should have been more clear it was only an interpretation, though.  The first answer there says it better.

The famous notion describes every proton and antiproton in the universe as a single particle moving backwards and forwards through time.

(i never bought it, but it's cool)

(that proton is a champ)

[Mika]

What do you know about my education? I've been a finalist in physics olympiads in my country, to which I had to learn quantum physics. Thermodynamics is ****ing easy in comparison.

You said yourself you are an architect. Participating in Physics olympics is good, but now I'd like to see your publications about quantum mechanics and quantum dynamics also.

Any question related to Arrow of Time usually tends to lead to the nature of time itself, hence my answer.

I think the way to start thinking about this is to gather what is known, Planck time is the shortest observable time by the current understanding. Beyond that, we cannot know whether time is stepwise or continuous.

[Flipside]

Let's be a little careful about E-Dick waving or making definite statements about an unknown quantity please, we can have a discussion without it

On subject, I think the problem is that 'Time' is kind of two different sciences. As Luis says, for a human "Time flies when you're having fun", but that's a perception thing. The life of a Star is still the life of a star.

I, for example, cannot image certain reactions taking place in reverse, because that not only implies that time is going backwards, but some major laws are being broken. I think that the concept of it being like a video on rewind probably wouldn't work and that Time is more complex than that.

[Ghostavo]

I've always thought that the increase of entropy marked the direction of where time flowed.

I remember reading that this made the Big Crunch theory somewhat weird.

[Luis Dias]

What do you know about my education? I've been a finalist in physics olympiads in my country, to which I had to learn quantum physics. Thermodynamics is ****ing easy in comparison.

You said yourself you are an architect. Participating in Physics olympics is good, but now I'd like to see your publications about quantum mechanics and quantum dynamics also.

Do you also ask for mathematics papers from your interlocutors if they teach you what matrixes are?

Don't believe me if you want. I've linked you to a good video that is not boring, and should be enough for you. The person in question has done something about quantum mechanics (mother of all euphemisms).

Any question related to Arrow of Time usually tends to lead to the nature of time itself, hence my answer.

Yes, but that's due to the nature of the conversation more than the nature of the question. People like to talk about metaphysics, and so they will ask things like "what is Time itself?", instead of focusing on the answers that were provided to the precisely worded question, that was about why there is an arrow of time at all. Which has a simple answer. QM et al may not be time symmetrical, but as I said, it's a can of very difficult worms that are not that relevant to upper scales of existence, so to speak. Entropy really is the key to understand why we do remember about the past and not the future, why time really seems assymetrical to us.

I think the way to start thinking about this is to gather what is known, Planck time is the shortest observable time by the current understanding. Beyond that, we cannot know whether time is stepwise or continuous.

Or if those categories are even useful. Mostly, they seem not to be. There are no hidden variables inside the quantum soap, which means that these questions are irrelevant.... and specially irrelevant towards the theme in question. You're trying to answer a question about the forest using solely a microscope. You're missing the big picture, where the stuff is actually happening.

[Luis Dias]

I've always thought that the increase of entropy marked the direction of where time flowed.

You thought correctly.

[General Battuta]

I've always thought that the increase of entropy marked the direction of where time flowed.

You thought correctly.

Not quite - or at least, not completely. The thermodynamic arrow of time is not an adequate explanation and does not even encompass all known arrows of time (though it gets most of them, including, in the modern synthesis, the quantum arrow of time).

There have also been a number of statements here that physics is not time-symmetric, but at the subatomic level this seems like far too strong an assertion for me - at least off the top of my head I believe most subatomic processes are time-symmetric. The obvious exceptions are (probably) some weak forces processes. I am particularly interested in the weak arrow of time as it seems to provide a possible (how well-substantiated?) explanation for the predominance of matter in the universe.

[Mika]

Do you also ask for mathematics papers from your interlocutors if they teach you what matrixes are?

Don't believe me if you want. I've linked you to a good video that is not boring, and should be enough for you. The person in question has done something about quantum mechanics (mother of all euphemisms).

I should have done that even more when I was in the University back then. If a teacher is smart, he's glad his student is curious.

By all indications, what you are doing here is repeating what Feynmann said without deeper understanding why things are the way they are. Feynmann has a lot of merits in different fields of Physics, including popularization of the quantum world. There's a catch though, by only reading the popularized texts one is still unable to predict what will happen once the question is not mentioned or covered by that particular text. The other thing to consider is that the image you get from the popularized text might sound reasonable, but could still be very much wrong - a convenient lie that makes understanding easier if you have the necessary background, or to give you an approximate mental image of what's going on if you don't. But there is still a possibility that you got it all wrong too! Hence I ask your publications on the field. I don't consider myself an expert in the field (with a possible exception of photons) and reserve that I might actually have it wrong, but I also give Feynmann the possibility of being wrong. Bear in mind, the way he describes some of the quantum interactions are how he understood them and the justification of continuing to use them is that they seem to work and describe the reality adequately.

I also have a (faint) recollection that Feynmann had a rather short temper on students who just parroted the results and really pushed his students to get over that.

Battuta, I get the feeling that you are also aiming at something else by this question. What would be your hypothesis on the weak arrow of time?

[General Battuta]

Battuta, I get the feeling that you are also aiming at something else by this question. What would be your hypothesis on the weak arrow of time?

Well I'm not really sure, I haven't kept up on it - but am I wrong in recalling that there are certain (apparent) time asymmetries in kaon decay that might have ramifications for cosmology?

[Astronomiya]

Well I'm not really sure, I haven't kept up on it - but am I wrong in recalling that there are certain (apparent) time asymmetries in kaon decay that might have ramifications for cosmology?

Nope.  Read on below.

There have also been a number of statements here that physics is not time-symmetric, but at the subatomic level this seems like far too strong an assertion for me - at least off the top of my head I believe most subatomic processes are time-symmetric. The obvious exceptions are (probably) some weak forces processes. I am particularly interested in the weak arrow of time as it seems to provide a possible (how well-substantiated?) explanation for the predominance of matter in the universe.

The electromagnetic and strong interactions are time-symmetric so far as we know, as is our only field theory of gravity (Newton).  However, GR is not, which leads me to believe that the eventual theory of quantum gravity we come up with won't be either (especially with Gravity Probe B just having confirmed both effects it was looking for to good precision).  As you noted, the weak interaction isn't time-symmetric either, due to its CP-violation.

CP-violation in the very early universe is almost certainly part of the explanation of baryogenesis; however, it is not the only necessary condition.  There are three of these, called the Sakharov Conditions:

1.  Baryon number must not be conserved.
2.  C- and CP-violation must occur. (Note that this is fully equivalent to saying T-violation must occur if CPT-symmetry holds)
3.  The interaction must not take place in thermal equilibrium.

What is interesting is that the strong interaction doesn't seem to cause CP-violation, seeing as there are terms that could easily do it in the QCD Lagrangian.  However, if these terms are not zero or extremely close to it, you would expect to see the neutron have an electric dipole moment about a billion times stronger than it is experimentally constrained to have.  The upshot of this is that CP-violation in the Standard Model is not nearly strong enough to cause the distribution of matter seen today (according to Wiki, using weak interaction CP-violation, you get enough baryons to make all of one galaxy).  So either strong CP-violation occurs at extremely large energies, or there's new physics there we don't understand yet.  It's a really neat question to think about.

[Herra Tohtori]

The electromagnetic and strong interactions are time-symmetric so far as we know, as is our only field theory of gravity (Newton).  However, GR is not, which leads me to believe that the eventual theory of quantum gravity we come up with won't be either (especially with Gravity Probe B just having confirmed both effects it was looking for to good precision).  As you noted, the weak interaction isn't time-symmetric either, due to its CP-violation.

How is electromagnetic interaction time-symmetric when the behaviour of photons (which are the bosons of electromagnetic interaction) is clearly not time-symmetric?

In fact I don't see any part of any interaction described in quantum mechanics to be time-symmetric due to the statistical nature of wave functions which is what quantum mechanics provides as far as predictive power is concerned (Schrödinger and Dirac equations mainly).

It's like... if you have a mathematical function f(x)=y with input set and output set. To be a valid function, every input value must result in one output value, but several input values can have the same output value.

However, the function can only have inverse function f^-1(y)=x if each input value has an individual output value.

If there are several input values that produce the same output value, then if you invert the function you end up in a situation where the inverse input set (original output set) has values that would need to produce several output values, and this means the inverse function is not a valid function.


Quantum mechanics sort of has a similar situation. With proper progression of time, you have input set of values that you feed into Schrödinger equation and if you feel particularly masochistic, Dirac equation - and they spit out a wave function that will give you a good idea where the particle will end up. The problem here is that the output is a wave function rather than a discrete value.

Now, it is possible (to some extent) to invert these functions - if you have the actual wave function that you know. However if you look at the situation from inverted time perspective, you'll see some values (with appropriate uncertainties) for a particle's momentum and location, rather than the full wave function of them.


So now it becomes impossible to construct the particle's original input values from its future values, because (to use Copenhagen interpretation terminology) the wave function has collapsed and doesn't exist any more. Knowing a particle's momentum and location makes it impossible to know the actual wave function the particle used to have. If you have several observations for a particle's values, then you can approximate a wave function, but a time-symmetric quantum theory should be able to do the impossible of extrapolating the wave function out of a single observation.

This becomes even more problematic when you consider that a time symmetric quantum theory should be able to, let's say, look at individual photons passing through a double slit in inverted time, coming in from different directions from the shader and converge in their original emitter location.


I understand you were talking about sub-atomic processes but there is no fundamental difference between sub-atomic and superatomic processes, just as there's no difference between microevolution and macroevolution.

As an example, electromagnetic interaction is transmitted via photons (virtual and real); therefore the problems with photon behaviour in inverted time are very much relevant to any process involving electromagnetic interaction...

And same problems would be evident in any quantum model which relies on wave functions rather than discrete predictions...

At least, that is my reasoning. If you can spot any obvious errors in my argumentation, please do point them out, as I'm fairly tired at the moment and ready to go to bed.


EDIT: Of course, how much this is apparent depends on the interaction in question. Z- and W-bosons are fairly massive particles, as is gluon; as a result, they may exhibit more particle-like tendencies than wave-like tendencies in most situations - but due to wave-particle duality, they would still have wave-like characteristics to some extent... and that sort of throws a wrench in the time-symmetry thing, as far as I'm concerned.

[Astronomiya]

Your thinking on the matter is not quite right, but I'm not entirely sure where you're going wrong right now.  Here's a quick proof that QED has T-symmetry:

By the CPT Theorem, all interactions in the universe have CPT symmetry.  Electromagnetism has both charge and parity symmetry.  That is, it behaves exactly the same way if we swapped the label of all charges (which should be obvious), and if we rotate by 180 degrees in ONE spatial dimension (this requires a little more thought to see; consider point charges first, then extend it to diffuse charge distributions via superposition).  Therefore, via the CPT Theorem, the EM interaction must have T-symmetry as well, for if it did not, it would break CPT-symmetry, which it does not.  QED

Also remember that while QED and QCD might be time symmetric, QFT as a whole is not.

[Herra Tohtori]

Your thinking on the matter is not quite right, but I'm not entirely sure where you're going wrong right now.  Here's a quick proof that QED has T-symmetry:

By the CPT Theorem, all interactions in the universe have CPT symmetry.  Electromagnetism has both charge and parity symmetry.  That is, it behaves exactly the same way if we swapped the label of all charges (which should be obvious), and if we rotate by 180 degrees in ONE spatial dimension (this requires a little more thought to see; consider point charges first, then extend it to diffuse charge distributions via superposition).  Therefore, via the CPT Theorem, the EM interaction must have T-symmetry as well, for if it did not, it would break CPT-symmetry, which it does not.  QED

Also remember that while QED and QCD might be time symmetric, QFT as a whole is not.

OK here's the problem I have with this, because two image tells more than two thousand words.



Here we have a classic double-slit experiment. Coherent light (or individual photons) are emitted from a light source, travel through double slit that is narrower than photon's wavelength, and an interference pattern emerges. The interference pattern can be characterized by a wave function which tells us the probability for different arrival locations on the shader. It's impossible to know exactly where individual photons emerge, but in statistical analysis the predictive power of the wave function is incredible.

Now, when we reverse the time we have the following situation:




We have the shader emitting photons into the direction of the double slit, and after they pass through the slit they... somehow should converge to their original location.

In the context of quantum mechnics, this would require a wave function with just one insanely high spike at the middle - probability of each photon arriving at their original location on the emitter should be 1, and I have a big problem with this expectation.

If we assume the model to be time symmetric, then we should expect the model to deal with this dilemma somehow and predict that all the photons coming from different directions will converge to the light source (which, in inverse time, acts as a photon absorber rather than emitter but eh, details). To be time-symmetric, a model needs to be able to be run from status A to status B, and from status B to status A in inverted time, but much like with inverse functions, you run into big problems when you have output values that correspond to multiple input values...

This is impossible and absurd expectation and, in my mind, it is irrefutable proof that quantum mechanics is fundamentally not time-symmetric. Photons are involved in everything where electromagnetic interactions are handled, and if photons can't be handled in a time-symmetric way in such a simple setting, then by the power of greyskull they can't be handled in a time-symmetric way in more complicated settings such as sub-atomic interactions. If anything, in smaller scale these quantum phenomena would be even more prominent.


Now, this is obviously not proof that the universe is not time-symmetric, although there are some pretty strong suggestions that statistical properties of quantum mechanics have it right and the universe indeed does have an element of randomness (which is impossible to reproduce if you invert time).

[T-LoW]

Professor Harald Lesch - What is time? (in german )

This man can explain everything. Could listen to him all day. Maybe someone knows a good translator?

[Mika]

Herr Doktor, think about the number of photons arriving on the shader first! Is the probability that all the emitted photons hit the shader actually 1?

[Herra Tohtori]

Herr Doktor, think about the number of photons arriving on the shader first! Is the probability that all the emitted photons hit the shader actually 1?

Technically that depends on the shader's structure - if it's flat shader, then no, but if it's a half sphere...

The wave function spreads to infinity, and when integrated over negative infinity to infinity, it results in probability of exactly 1 (meaning that all the photons go somewhere), but that is really irrelevant for the issue at hand.

The problem is that all those photons that go through the double slit (regardless of where they end up at in real-time) will come back in inverse time and they WILL converge upon the exact point where their journey began. If the inverse time model predicts a different spot for the photons - or, heaven forbid, a wave function, then clearly the model isn't time symmetric in the sense that I understand the term.

And that's the problem here - as far as I'm aware, truly time-symmetric model would require discrete predictive power with exact solution for an exact experiment, rather than statistical prediction of how multiple repeats of the experiment will happen. Quantum mechanics doesn't provide us with that sort of predictions.

[Mika]

And the absorption by the slit?

[Herra Tohtori]

And the absorption by the slit?

...every photon that goes through the slit, goes through the slit.

This is a tautology.

In inverse time, all those photons come back through the slit from the other direction.


The rest of the photons emitted by the emitter are irrelevant. We can run the experiment with single photons if you'd like, and in inverted time, each of those photons will be reverse-emitted from the location where they ended up, travel to the slit, change their course to align with the original emitter, and impact the emitter (visibly changing the energy state of some electron in the emitter source).

If the theory can't predict that this will happen, then it ain't time symmetric. And the best quantum mechanics offers are statistical predictions, so it seems to me that time symmetry in context of QM should just be abandoned as a bad job.

[Mika]

I'm not sure if I understood what's the problem then.

However, tracing photons from the emitter to the shader loses some part of light on the slit. The setup is not a true reverse system if you don't include those lost on the slit itself (at exact phase, location, direction, time and polarization) when going through it backwards.

Could you explain why a delta spike as a wave function disproves the whole reverse thing?