Redingold said:
This end of this video is no longer in agreement with scientific study. Last year, a team of physicists doing a double slit experiment managed to observe the path of a particle, noted precisely which slit it went through, and still were able to produce an interference pattern. The point is that observation is not a passive process on the quantum level. In order to observer the location of, say, an electron, you need to actually do something to it - bounce a photon off it, for instance. Doing this changes the behaviour of a particle, but the reason is not so mysterious - a particle being struck by photons will obviously behave differently than one not being struck. If you're very clever with your measuring, then you can reduce the effects on the particle while still knowing where it is (to within the tolerances permitted by Heisenberg).
http://arstechnica.com/science/2012/05/disentangling-the-wave-particle-duality-in-the-double-slit-experiment/
People tend not to get wave-particle duality. It isn't that things are sometimes particles, and sometimes waves. Quantum objects are particles, and they propagate in wave-like ways. A particle has a thing called a wavefunction associated with it. The wavefunction behaves like a wave, and the position of the particle is related to the strength of the wave - more precisely, the modulus squared of the wavefunction gives the probability density function of the position of the particle. Interactions with other particles are...a whole lot more complicated, and not something on which I have much technical knowledge. Still, there's nothing mysterious going on, except for entanglement, which is freaking crazy.
I don't really know what to take from it. Reading that article, it sounds like what they've found is that you can observe a photon's trajectory without collapsing the wavefunction of other photons entagled with it, which I feel doesn't necessarily have as much to do with wave-particle duality as it has with entanglement.
The way I see it is that particles are expression of a wave, i.e., waves that interact with the world as points/particles. That is to say, a photon being fired at a detector is going to hit the detector at a random point, any point (within the extent of the wavefunction of course). However, not all points are equally probable for the photon to end up at. The probability of the photon being at a certain point is represented by the wavefunction. The point that the photon ends up at is still random, but it's like a dice where 3 of the sides say "5": the outcome of the die-roll is random, but some outcomes are more likely than others.
The waveform interacts with the world around it like other waves, and as such interference patterns appear (in the wave which is a description of the probability that the photon will be detected at a given point) when the waveform passes through two slits, creating these hotspots where the probability of the particle appearing if it is observed is extra high.
Imagine placing a row of detectors which detect photons but do not stop them along the path of the photon being fired.
This would allow us to "see" the particle travel in a "trajectory" (That is to say we would end up with a set of points in space where we know the photon appeared, then we can draw a line through them and call it a trajectory[sub] However, I feel that drawing that line is completely meaningless and simply a flaw in the way we want to look at the particle[/sub]). The trajectory could be zigzagging and snaking about completely willy nilly, because the particle is appearing at a random point each time a detector collapses the wavefunction so that any trajectory is actually possible. That is to say the photon, which I consider to be a wave, has to "Be considered/become/appear as/spawn" a particle to interact with each detector and the spawning of this particle happens on a random - albeit affected by probability - point on each detector.
What the experiment you've linked me to seems to imply in my mind is as follows:
If you fire two entangled photons at two 100% identical detectors, the two resulting points of impact will be on the same point on each detector will. This holds true regardless of where along the path of the photons you place the detectors, so the photons are following the same imagined/hypothetical paths. In other words, the wave's random spawning of a particle for interacting with a detector is identical for two entangled photons. The waves spawn particles at the same points because they are entangled. However, collapsing the wavefunction of one photon will not collapse the wavefunction of the other. This means you can place detectors in the path of one photon and a double slit in the path of the other and then with the help of the one photon see which slit the photon passing through the double slits would be passing through had you placed a detector by the slit. But even though this imagined trajectory is tracing through one of the slits, I don't think it makes sense to say that the photon is passing through the slit in question: its waveform is not collapsed, and the wave is therefore passing through the slits as a wave would (or the particle is propagating through the slits like a wave, if you like). The other photon is merely marking one point on the hypothetical path the particle would have appeared on had you been observing the wave and collapsing it.
Say you placed a double slit in the path of each photon.
Photon 1's double slit has a detector in one slit, letting us know which slit it passed through. Photon 2's double slit has no detectors. There are of course detectors placed behind each set of slits.
(This is pretty much analogous to what the experiment you linked is doing, except here we are examining where Photon 1 ends up after being observed at the point where Photon 2 passes through the slits and we're giving both of them slits in order to make their situations identical save for the detectors.)
I dare say Photon 2 would produce an interference pattern, while Photon 1 would not.
I say this because Photon 1's wavefunction is being collapsed, while Photon 2's is not.
Since Photon 1's wavefunction was collapsed when it was observed by the slit, I'm thinking its movements from there on are described by a wave that originates in that slit, creating that two-band particles-through-slits effect. Perhaps after each observation the wave is set up anew meaning that Photon 2 is described by the same wave it started out as with all the twisting and turning space around it (most notably the slits) is imposing on it. Meanwhile, Photon 1 is observed passing through the slit, and since we know its position at that point its movements from then on are described by a new wave that passes through only one of the slits at a time so it does not experience interference.
Back to the original experiment:
The way I see it, Photon 1, being observed, shows us the point at which Photon 2 would have appeared if we'd observed it. However, with Photon 2 passing through double slits, the wavefunction - and with it the hypothetical trajectory - changes. It changes gradually though, so if you place Photon 2's detector close to the slits, Photon 2 will be observed as coming out of the slit Photon 1 indicates it would have passed through since it's still continuing from that hypothetical trajectory. However, Photon 2's wavefunction has not been collapsed, and as such its movements must still be described by the original wavefunction, and this wavefunction is being bent by the slits its passing through. If you place Photon 2's detector far enough away from the slits for the wave to interfere with itself, the photon will behave so as to create an interference pattern even though we observed which slit it would have passed through. Photon 2's hypothetical trajectory (or rather set of trajectories, since we're firing several photons in succession to produce the interference pattern) is now changed to one that will produce an interference pattern; something that Photon 1's trajectories will not since they're being restarted at the slit each time it passes through.
I've also been thinking that it's possible that placing the detector as close to the backside of the slits as they did to confirm that Photon 2 passes through the slit Photon 1 says it does, is actually equivalent to placing a detector in the slit. After all, the effect is the same: you get to know which slit the photon passed through.
So I think that saying that the photon is actually going through one of the slits when it isn't being observed in the double slit experiment is faulty, because that photon isn't acting as a particle at that point.
I'd say it makes more sense to say that it
would pass through *this* slit if we observed it, rather than saying that it does pass through *this* slit.
I think the photon has the property of a hypothetical/imagined trajectory passing through one of the slits, but saying that it actually does pass through one of them would imply it expresses itself as a particle, collapsing the wavefunction and leaving us with no interference pattern.
I'm sorry if this makes no sense at all. It's the middle of the night and haven't had the time to read over it properly.