There is a story in the popular science press about the Large Hadron Collider and a time-travelling Higgs boson. It was done, I am sure, with tongue firmly in cheek, but the idea is based on some interesting ideas in modern physics.
I was going to make an attempt to explain what they were talking about, but I think I may need to work up to it gradually. The difficulty is that one gets distracted by the explanation of the physics from the mind-blowing implications of the conclusion. So I’ll talk first about something a little easier, that is still pretty interesting.
What I’m going to do is show you how to catch a burglar with a light-beam alarm down which no light passes. This ensures that even a super-sensitive burglar who can detect single photons hitting him may be totally unaware of having tripped the alarm. It will also demonstrate that things that might have happened but didn’t have real, observable effects.
Here is the basic set-up:
On the left we have a light source that hits a sheet (A) of half-silvered glass angled at 45 degrees. Half of the beam is reflected, and half passes straight through. There is nothing magical about this – it is the same piece of physics as when you see your reflection when looking out through a window at night.
The two beams split up and are reflected off mirrors at B and C. They bounce back to a second half-silvered glass just like the first. Again, each beam can either pass straight through or be reflected. But if you get all the distances exactly right, then something slightly odd happens. All the light comes out at D, and none at E.
There is an easy explanation for this if we think of light as a wave. The two waves going out through D line up, and reinforce one another. But the two going out through E cancel out, peak-to-trough.
But things get harder to explain when we turn the light intensity down so that we have only one photon going through at a time. Now any one photon on hitting the half-silvered glass either reflects or is transmitted at random. If you put detectors on either side, only one of them ever goes off at a time. And when it gets to the second half-silvered glass, again it can reflect or transmit at random again. But if it reflected at the first glass, it will always transmit at the second, and vice versa. So the photon always comes out at D, never E.
So far, so strange. But now it gets really weird. Now we place an obstruction in the light’s path between A and B. The photon can go either way at A, and if it reflects it will get blocked. But if it goes via C, it reaches the second glass, and can now go either way, and it really can! Now half of the photons come out at E!
Now we know that any photon that appears at E must have gone round the ACE route, and therefore never went anywhere near the obstruction. Even if the obstructor were able to detect single photons, it would have seen nothing. But the fact that it came out at E tells us that the obstruction is there.
So we can use this as our burglar alarm. The path from A to B is set across the doorway. The burglar acts as the obstruction. And even though our photon never went anywhere near the burglar, we can still see him!
The philosophy of counterfactuals – things that could have happened but didn’t – has always been difficult. How can you study events that never happen?! But in quantum mechanics, they’re real. Because the photon could have been blocked but wasn’t, it can come out at E. If the photon could have gone either way, it will always come out at D. Strange, but true.
-
Next up: the Delayed Choice Quantum Eraser. This is where we start messing about with time.
Great stuff.
Let’s see if I have this correct :-
If the mirror at B was removed 50% of the photons would then begin to hit the obstruction and 25% arrive at D and the other 25% at E ?
In other words the ‘potential’ of an ABD path, even though it’s blocked, forces a photon to transmit rather than reflect at A ?
Yes, that’s right. Though it would be more correct to say it enabled it to transmit at the second half-silvered glass.
There’s even a variant I’ve seen where you partially block the path, by putting some star-shaped (say) obstruction in. Out of E you will see a star-shaped patch of light, appearing where the obstruction blocks its passage.
Well, IMHO, the thing about a scientific theory is that it should explains the observations. Explaining what has been theorised but not observed is not a requirement.
If one considers that the wave theory of light and the corpuscular theory apply all the time, even when no observations are being made, there is a problem. However, if (in this experimental setup) one considers that the corpuscular theory only applies if and at the point where the setup includes a corpuscle (photon) detector, all becomes OK.
If one wants, one can view it that half photons do (in fact any fraction of a photon does) propagate, because the wave theory applies no matter what the beam intensity; the quantisation into photons (with the necessary non-zero probability of the photon not being there) only comes into effect when an attempt is made to detect the photon (necessarily as a corpuscle, I think) by its reaction with things other than a light beam at the same wavelength (eg electronic vacuum tube or semiconductor photon detector, photographic plate of atoms chemically changed by photon absorption).
Concerning the burglar alarm scenario: surely if the burglar has a photon detector, he could detect, in the path AB, half of the total number of transmitted photons? But, of course, each detection (some of which are presumably when he does not need to be crossing the light path) will risk triggering the alarm: but surely only for 50% of the emissions. So, he cannot reliably detect the light path unobserved, but he can detect it and there is a 50% chance for each of his detections that it will be unobserved. I hope that’s right; it’s certainly complicated.
Best regards
I shall come on to whether you can resolve things at the time of measurement shortly. Briefly, yes you can, but there are some issues that make even this tricky. I prefer a different way of looking at things, but it’s still a subject of considerable controversy.
Yes, the burglar has a 50% chance of seeing the photon, and a 25% chance of not being detected because the photon came out at D, but if a single photon is passed through the system and comes out at E, he gets detected by a photon that (arguably) didn’t go near him. That this can happen at all is what’s interesting.
I know I read somewhere - I believe it came from “Anglican Samizdat” - that if this trick is done with optical fibres, and the ends are far enough apart, the reaction when one is affected is mirrored in the other - even though the distance means any communication by conventional means is 10,000 times faster than light.
IMHO, that means instantaneous.
But it was intriguing; if I can find the link I’ll post it.
Found the link.
http://discovermagazine.com/2009/may/01-the-biocentric-universe-life-creates-time-space-cosmos/article_view?b_start:int=0&-C
Though i don’t buy the “biocentric universe” idea.
El Draque,
That’s referring to an example of the Einstein-Rosen-Podolsky (EPR) experiment. Certain atoms can be induced to emit two photons at once, which in order to make all the forces balance must have the opposite polarisation (so the total is zero). The thing is, though, the polarisation of either photon alone is completely random.
So the problem is, when is the random choice made? When they split? Or when they are measured?
If it is when they split, then it is like the burglar alarm situation above in that there still seems to be some echo of the other alternatives remaining. And in any case, this situation would lead to measurements made in other ways giving results that real photons in fact do not. They’ve done experiments to rule this case out.
On the other hand, if it is when they are measured, then the random choice made at one measurement apparently has to be conveyed faster than light to the other experiment so that it will come out the right way. And because of relativity, this effect has to be able to travel both instantaneously and backwards in time, too. While it doesn’t allow FTL communication, or break any other rules, and more importantly it is confirmed by experiment, it is at the same time somewhat upsetting for physicists.
There is in fact a third option - that the random choice is never made. In the same way that the photon can travel both paths at once, the experiment can produce both outputs at once, and when the experimental results are brought back together at the other end, the waves interfere and any inconsistent results cancel. They mysteriously never happen, in the same way the recombined photon never comes out at point E. For some reason, a lot of physicists don’t like that answer either. But it is in many ways the simplest.