= 27d567bc2d48d9f40e9ccc20ea370b15)
timtyler comments on Quantum Physics, CERN and Hawking radiation - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (66)
Black holes not being able to lose mass would violate reversibility. If something can fall into a black hole, other things must be able to come out again - since physics exhibits microscopic reversibility, and the known laws of physics are symmetrical under T=-T.
This is probably not correct. We know that CP symmetry is broken. There are good theoretical reasons to believe that CPT symmetry holds. Consequently T symmetry must be broken.
To quote from here:
...for the reason to think that it is T symmetry, not CPT symmetry that holds.
Charge and Parity are likely to be implemented using internal rotation or sequences - and so will reverse automatically if T=>-T. We don't have to manually reverse particle momenta or spin if we are running things backwards. That happens automatically. The idea is that Charge and Parity function like that too - due to how these phenomena work. For instance, Charge seems likely to be implemented rather like a bi-directional pump. Reverse time, and such a pump runs backwards automatically. You do not have to go around reversing all the charges - they reverse themselves automatically. It works like spin does - and reverses automatically for much the same reason.
T-symmetry is neater and simpler. We should prefer it - unless there are good reasons not to do so.
If so, then how can T symmetry hold? You seem to be saying that T symmetry implies CPT symmetry. But we know from experiment that CP symmetry is broken. If T symmetry holds, and CP symmetry does not hold, then CPT symmetry cannot hold.
Really, this looks pretty straightforward. The theory you quote has A->B. Experiment !B. Consequently, either !A or !(A->B).
Why do you think so?
Particle momenta, no; spin, yes. Although spin is angular momentum, it does not come about because particles are rotating about an internal axis, as you seem to have in mind. (To the best of anyone's knowledge, of course.) Consequently parity does not auto-reverse under time-reversal.
OK - so, you don't understand the idea. There is a much more detailed description of the associated model written by someone else here.
The punchline at the bottom reads:
Please let me know if that fails to sort you out - and you are still interested.
First, the theory rests on the airy assertion that reversing T automatically causes the reversal of spin and other quantum numbers as well. I found the argument given for this unconvincing. Second, and more importantly, you do not seem to have grasped that you cannot possibly have both T symmetry and CPT symmetry, because CP symmetry is experimentally excluded. It does not matter if you invent a special form of T symmetry that is 'equivalent' to CPT symmetry.
Take a physical system that exhibits CP violation; assume it is described by the kind of theory outlined in your link. Now reverse time. By the argument in your link, this also reverses CP. Because the system is not symmetric under CP, it exhibits different behaviour. Bing, T symmetry has been broken: There is a measurement I can make that tells me which way time is flowing.
Well, I don't have a watertight argument for the first point. I think it is more likely than not, but if your intuition is the other way around, I won't argue too much. What I object to is the idea that T-symmetry is wrong. In fact, T-symmetry is pretty plausible, IMO.
From your second point, (from my perspective) you still don't get the logic of the whole idea - and you have exhausted most of my resources on the subject, so I am not sure what more to do with you.
Assuming that charge and parity quanta involve moving parts internally, then they would both reverse automatically if time is reversed - producing what appears to be CPT symmetry as a result. That would be consistent with all known experiments, and physics would then by time symmetric.
You said: "Because the system is not symmetric under CP, it exhibits different behaviour." No, because you have also reversed time, (you just said so yourself) - and if C,P and T are all reversed, then symmetry is restored. So, then there is no measurement you can make that tells you which way time is flowing.
No. Start with a left-handed neutrino. Reverse T under your assumption. It is now a right-handed antineutrino going the other way; reverse space as well to restore the original direction, if you like, although the argument does not depend on this. Because CP is broken, right-handed antineutrinos do not behave exactly as left-handed neutrinos do. Therefore you can tell how many times T has been reversed. You don't get the full symmetry back except by applying CP another time.
Yes.
A parity flip, I presume you mean.
That is indeed true.
Well you only said you reversed it once - and then you flipped P, but not C, leaving things in a bit of a mess - and then you tried to make out the mess was something to do with me.
Reversing T an odd number of times changes everything. Reversing it an even number of times changes nothing. You can't distinguish between reversing T different numbers of times beyond that - under the hypothesis that reversing T automatically reverses C and P.
Ok, leave the parity flip out of it. If this is true:
then you do not have T symmetry. Done.
So, which of the hypotheses of the CPT theorem do you find less compelling than, er, the fact that it would be kinda neat if T symmetry were correct? (If there's any reason beyond that in the page you linked to, I failed to see it.)
Well, T symmetry is favoured by Occam's razor. We have pruned away the momentum quanta from needing to be replaced by their anti-particles. The idea that T-symmetry is true just takes this a bit further, becoming more elegant and neat in the process.
You didn't answer the question, nor did you give any grounds for thinking it doesn't need answering.
Occam's razor applies to theories, not to individual propositions about them. CPT (or T) symmetry isn't something you build into a physical theory by having an axiom like "CPT-symmetry holds"; it arises from the structure of the theory. Do you have any actual reason for believing that theories with T-symmetry but not CPT-symmetry are simpler than theories with CPT-symmetry but not T-symmetry? The CPT theorem seems to me to give good reason not to believe that.
Well, I think so, but maybe not in a format suitable for a short blog post. There are numerous small, simple CA with the property of being symmetrical under T=-T. For example, the BBM. My impression is that other means of reversal are correlated with automaton complexity. Then there's the idea of charge as a pump. That is appealing on other grounds - and pumps tend to have moving parts - which would then reverse automatically if T=>-T . Also, the possibility of simulationverse ideas would seem to favour ease of reversal, to some extent.
IMO, you should really not be counting the CPT theorem as evidence on the issue - one way or the other.
I certainly don't think people should be telling me that the known laws of physics are not symmetrical under T=>-T. IMO, it is more probable that they are symmetrical that way than that they are not. The idea that CPT symmetry illustrates that they are not is simply a popular misconception, with no basis in the facts of the matter.
The laws of physics as currently understood -- i.e., the laws in the best model we' ve got -- are in fact CPT-symmetric but not T-symmetric. (Because the best model we've got is a quantum field theory of the sort that the CPT theorem applies to; and because CP symmetry is violated (1) by that model and (2) in reality, according to the available evidence.)
Sure, there are plenty of small simple cellular automata with T-symmetry. And also with P-symmetry, which does not hold in the real world. So far as I know, CAs with PT symmetry are just about exactly as easy to make as ones with T symmetry. (And if you have CAs with a property corresponding to C, I bet CPT is as easy to arrange as T.) Why is any of this meant to mean that T-symmetry is simpler than CPT-symmetry?
You may find the idea of "charge as a pump" appealing; fair enough. I am at a loss to see why that is a reason for thinking that T symmetry is simpler than CPT symmetry.
Your argument, so far as you've provided one, seems to go like this: "I expect the universe to work like a cellular automaton. I have the feeling that T symmetry is simpler than CPT symmetry in cellular automata. Therefore the CPT theorem is irrelevant and we should expect T symmetry to prevail but not CPT symmetry". This strikes me as very strange since (1) we have very successful physical theories that are QFTs (to which the CPT theorem applies) and no successful physical theories based on cellular automata, and (2) the CPT theorem is an actual theorem governing QFTs, whereas your intuitions for relative simplicity are merely your intuitions.
You understand that the claim is that that is just a historical accident about the way the model was built? The idea that C and P reverse automatically if T is reversed does not make any new predictions that the old model did not. The idea that CPT symmetry is favoured by some kind of experimental evidence seems completely wrong to me. Since the two models are totally equivalent experimentally, this is an issue for Occam.
There is a mathematical theorem that says that no model of the same kind can fail to have CPT symmetry. There are robust experimental results that say that the real world doesn't have CP symmetry. Therefore, it is not an accident that given the general kind of model it is it has CPT symmetry but not T symmetry.
If you are claiming that it's a historical accident that we have a QFT model and not a CA-type model, then show your evidence. Specifically, the obvious alternative hypotheses are things like "There is no CA-type model that actually fits the data" and "All CA-type models that fit the data as well as the current QFT model are much more complicated than the QFT model". Given the present state of the art as I understand it, these seem much more likely to me than your hypothesis that it's just a historical accident. Please feel free to convince me otherwise.
What two models? I know of one model, the Standard Model, which fits the data extremely well and is reasonably simple (ha!), and which has CPT symmetry but not T symmetry (and cannot be fudged to have T symmetry without losing its agreement with experiment). What is the other model you propose, that supposedly is totally equivalent experimentally but has T symmetry?
Please be specific about this, because either you have some wonderful physics to show me that I haven't seen before or you're failing to distinguish between "some vague intuitions in Tim's head" and "a very thoroughly worked out and tested physical model".
Well, if true, it is a reason to think T symmetry or PT symmetry holds - since then charge would reverse itself automatically if T was reversed.
Depending on the nature of the pump. For instance, a pump made out of Standard Model stuff would not exactly reverse itself if T were reversed. Presumably, then, you have something more specific in mind; would you care to say more exactly what it is?
That a time-inverted process doesn't contradict the fundamental laws doesn't mean that we could observe it with frequency basically comparable to that of the uninverted process. Think of thermodynamical irreversibility, for an example of class of processes which practically don't have inverted counterparts, even if these are perfectly compatible with microscopic physics.
Also, losing mass is not clearly defined process. If by mass of a BH one means the mass included below the horizon (or calculated from the diameter of the horizon) observed from constant distance, then black holes never lose or gain mass, since such an observer would never see anything cross the horizon in any direction: for an outside observer it takes infinite time for any falling object to reach the horizon.
If mass is calculated from the gravitational force measured in a constant distance, it may grow as mass gets attracted towards the horizon, or it may shrink if the mass near the horizon has enough outward momentum to overcome the gravity of the black hole. The latter scenario is quite unlikely to happen during typical black hole formation, at least I think so.