Yes, it’s from a Sci Fi magazine … and I had to dredge it up from the WayBack Machine (I started from Wikipedia’s page on the Afshar Experiment) but if I Google John G. Cramer I see that in addition to Sci Fi author, he’s Professor Emeritus of Physics!
What do the Board’s physicists think? Retrocausality is the answer after all? The Wikipedia page does summarize the controversy for sophomores. Any physicists here smart enough to summarize it for a Freshman?
I read it…but I confess the conclusion kinda escapes me. I was dubious about the bit where he predicts that “Quantum Lawyers” will have significance; I’m old-fashioned enough to believe that scientists will still be the ones driving the bus.
(On the other hand, there may be issues here that science cannot determine – in which case no one should be saying, “This is how it is,” only, “This is how it might be.” This is the position of String Theory. It isn’t “scientific” exactly, but it is perfectly valid to speculate upon.)
I would note that the transactional interpretation is John Cramer’s own interpretation, which (I believe) has been around since the '80s, long before the Afshar experiment. My impression is that it’s a minority view, but not a crazy one (at least no more so than other interpretations of quantum mechanics, and less so than the really crazy ones). He sees support for his interpretation in the Afshar experiment, but the interpretation of the Afshar experiment is (I gather) still controversial, so I’m not sure that changes much.
I don’t think most practicing physicists spend too much time mulling over what the correct interpretation of quantum mechanics is, although some do. And some would say that if all interpretations make the same predictions then it’s a philosophy question, not a physical one. I would speculate that those who are inclined to care about which interpretation is “reality” would be less likely to be satisfied with the Copenhagen interpretation, because the Copenhagen interpretation doesn’t really answer those questions so much as declare that they don’t matter. It’s less a well-defined interpretation than it is a miscellaneous assortment of things that Bohr and Heisenberg said. It has the virtue that it’s the most famous interpretation, and also that unlike the second-most famous (many worlds) it doesn’t require you to believe that a bunch of unobservable stuff exists.
Actually, after half-a-dozen or so semesters on quantum theory in college and grad school I’m still not sure I could say with certainty what the “Copenhagen interpretation” says – so far as I know, there isn’t one definitive version of what it says. But roughly:
(*There’s some business about how when you measure only part of a system you don’t see a pure quantum state, so if you wanted to see the state evolving unitarily you’d have to measure the combined state of the original system + measurement apparatus + the environment that the measurement apparatus interacted with – but I don’t think those ideas are really part of the Copenhagen interpretation. If any thing, they suggest that the whole universe – at least the portion of it that has interacted, directly or indirectly, with your measurement apparatus – really can be in a giant, unobservable superposition.)
At any rate, I don’t mean to be too hard on the Copenhagen interpretation. Most interpretations seem to either come down to: (1) Assume a bunch of stuff exists (e.g., other “worlds”) that you can’t ever observe, even in theory, (2) assume a bunch of stuff doesn’t exist (e.g., the wavefunction) but nevertheless is somehow useful for making really good predictions, (3) assume a minimal set of things really do exist (e.g., the wavefunction), but that the rules by which they evolve depend on vague distinctions like whether something is a measurement or not.
All of which are far better than (4) Let’s say consciousness has something to do with it, because we weren’t already confused enough!
He’s using “lawyers” in a derogatory sense, to mean scientists who will use bullshit arguments to continue to defend those interpretations he considers falsified.
I will say, I don’t totally understand this part:
(bolding mine)
If that’s the case, shouldn’t one be able to show this without relying on any experiment at all?
I think I’ll need to read up more on the various interpretations of the Afshar experiment if I want to make more sense of this.
Having thought about this a bit more, I’d say that if someone understands the Copenhagen interpretation to be saying something other than what the actual mathematical formalism of quantum mechanics says, then they’re understanding the Copenhagen interpretation in a non-standard way, because the standard thing to do (whether one claims to accept the Copenhagen interpretation or not) is to just do the math. People use the Copenhagen interpretation as some (rather vague) words to explain the usual calculations, not as instructions to do some different calculations. The reason a competent professional physicist might disagree with this is because, as I said, what the Copenhagen interpretation actually says isn’t very clear.
At any rate, regarding the Afshar experiment, I think it’s a mistake to think of the measurement of which hole the particle went through, which happened after the fact, as changing the past and having caused the particle to go through one hole or the other. Measurements don’t change the past, they collapse the state at the time of the measurement. (Even delayed choice experiments can be explained without retrocausality.) You might collapse the state to one that is the same as what you’d have gotten if the particle only went through one hole or only went through the other, but that doesn’t contradict the fact that, at the time before your measurement, it was in a superposition of “went through one hole” and “went through the other”. In this case, we know it was in a superposition, because it was prepared that way, and we didn’t do anything to collapse the state at that time.
Consider a simpler example. Suppose you prepare some light that’s polarized at 45 degrees. Then you pass it through a polarization filter that’s at 45 degrees. The filter won’t block any of the light, because the light is already polarized in that direction. Then suppose you measure the polarization of each photon emerging from the filter, in terms of whether it’s vertical or horizontal. Each time, you’ll find one or the other. Then you ask, if each photon was either polarized vertically or polarized horizontally, then why didn’t the 45 degree filter reduce the intensity at all? The answer is of course that they weren’t polarized vertically or horizontally until you measured the polarization – prior to that, they were polarized at 45 degrees.
I don’t really see why the Afshar experiment is any different. Sure, sometimes people talk about a measurement revealing the position of the particle – and Bohr and Heisenberg may have talked like this quite a bit way back when the theory was first being developed – but nowadays people don’t really mean it. You don’t just find out where the particle is by measuring it, you cause it to assume a definite position. People say things like “the particle has a definite position or definite momentum, but these can’t be measured at the same time”, but what they really mean is “the particle is in general in a state that has neither definite position nor definite momentum, but we can collapse it to a state where one or the other is definite, depending on our choice of measurement.”
I don’t think that’s quite the claim being made. Rather, the supposed ‘problem’ with the Copenhagen Interpretation as applied to the Afshar experiment is that you obtain both information about which path the light took—which hole it went through—and you see an interference pattern. So it’s not that the measurement retroactively caused the photon to have taken the path through a particular hole, but rather, that it simply reveals which hole it went through—which is then alleged, in combination with the existence of the interference pattern, to violate the CI.
Although what’s actually being ‘violated’ here is the principle of complementarity in particular, and in even more particular particular, the complementarity between the wave- and particle-pictures of light, with the ‘which-path’ information revealing the particle character, and the interference pattern revealing the wave character. I’m not sure, by the way, how this, if true, is consistent with quantum mechanics itself—after all, the relation between the visibility of the interference fringe and the distinguishability of the paths follows straightforwardly from the mathematics, without assuming any particular interpretation at all. Contrarily, if there is some reason for this relation to be violated in this experiment, then for the same reason, complementarity wouldn’t apply.
And I think you’re pretty much exactly on track regarding the reason why one can’t just naively apply the wave-particle complementarity relation here: the alleged which-path information is obtained simply in a second measurement after the interference pattern had been obtained; consequently, it’s simply no longer right to conclude that if I see a click at detector 1, the photon must have come from hole 1, too—i.e. the detection doesn’t actually give us any which-path information regarding the photon. In a sense, the claim is like saying that by first measuring the position, and then the momentum of a particle, due to getting definite values in each case, I have violated the uncertainty principle.
So I don’t really see any threat to the Copenhagen Interpretation (whatever one takes that to be, exactly) at all.
(Also, it’s somewhat amusing to notice, via the wikipedia page, that Ruth Kastner—who’s more or less the only exponent of the Transactional Interpretation besides Cramer himself—is in complete disagreement with the latter regarding the interpretation of the Afshar experiment.)
You might be interested in learning about Pilot Wave Theory, or the de Broglie-Bohm theory, which provides yet another non-Copenhagen interpretation of quantum behavior.
Here’s a recent article by Steven Weinberg, a Nobel Prize winner in physics, who explains why he has problems with any interpretation of quantum mechanics:
Famous experiments with “paradoxical” outcomes, e.g. Bell’s version of E-P-R or the GHZ experiment were predicted in advance, right? G, H and Z knew from quantum theory that they would achieve their paradoxical result once they got their experimental setup working properly.
Is Afshar experiment in the same category? If so, why did the experiment need to be performed before the controversy developed? Or was there some controversy (when to collapse a wave function? ) that Afshar resolved?
For me, it seems almost trivial to “explain” the EPR and GHZ paradoxes using something like Cramer’s Interpretation — quantum weirdness disappears! (Or at least is replaced with a very different weirdness.) But, even setting aside Afshar, is there a similar simple layman’s way to understand how Cramer (or retrocausality) copes with “paradoxes” of two-slit interference patterns?
The result of the Afshar experiment certainly follows from standard quantum mechanics, so it’s just a matter of doing the calculation right to predict what will happen in advance. And while I don’t know the actual historical order of events, it’s basically inconceivable that those calculations weren’t done beforehand, so I’m quite confident that the result was expected.
But of course, that’s no reason not to do the experiment. Take Bell-type experiments: the expected outcome according to QM was well known, but it conflicts with expectations generated by following some apparently physically sensible assumptions—locality and realism. So, in this case, it was actually shown that all interpretations of quantum mechanics that aim to explain it in a local realistic manner yield predictions at variance with standard QM; and of course, there’s then an a priori question of whether nature follows the quantum predictions, or those of local realism. And the definite answer to this question only came 2015, with the first loophole-free Bell tests being performed (and QM won out).
So it might be the case with the Afshar experiment that complementarity leads one to expect one result, while QM predicts another; and thus, the experiment might adjudicate between the two, just as it did with QM and local realism. However, I see no reason to believe that complementarity predicts anything that’s in conflict with the Afshar result.
Yes, that’s how things always seem to end up: you have to accept some weirdness; but you might get to pick and choose your favorite weirdness. Personally, I find retrocausality much more difficult to accept than the idea that not all events can be jointly ascribed values, and hence, don’t have a joint probability distribution (which is really all the Bell/GHZ-theorems say, mathematically), but I think reasonable persons may differ on that.
Thanks for your take on this. I think we are essentially in agreement. When I objected to people “thinking the measurement changed the past”, I think what I was trying to say was not that they were explicitly claiming that, but rather that it didn’t make sense to talk as if the wavefunction of the photon was already collapsed to “passed through hole A” or “passed through hole B” by a measurement that didn’t happen until later.
I agree with what you say here:
I’d written the following before noticing your link to Ruth Kastner’s comments, which include a similar argument:
Forgetting about the two-slit experiment for a second: If you have a spin 1/2 particle, and you measure whether it’s spin-up or spin-down in the x-direction, you’ll get one result or the other. By forcing its spin to assume a definite orientation in the x-direction, it must be a superposition of spin-up and spin-down in the z-direction. But then when you measure whether it’s spin-up or spin-down in the z-direction, you again get one or the other. It’s not that the measurement revealed that it was spin-up or spin-down in the z-direction all along – we know that at the time of the previous measurement, it wasn’t 100% either one – rather, the measurement changed the state to one in which the question “Is it spin-up or spin-down in the z-direction?” now has a definite answer.
) that Afshar resolved?