Insights Interview with Astrophysicist: Adam Becker - Comments

I just finished chapter 11 where Adam defends the various many-worlds views (string theory’s landscapes, inflation’s multiverse, and Everett’s Many-Worlds Interpretation, MWI). He admits MWI has a problem with the meaning of probability, but dismisses it as something to be solved in the future. I’m less optimistic, since the idea has been in vogue (in FoP anyway) for many years and yet the problem persists. For example, it can’t be simply that the branches split with a “frequentist interpretation of probability,” as Adam illustrates with the Schrodinger Cat in a 25% dead — 75% alive probability when there are only two possible outcomes. Another problem with a frequentist-splitting interpretation would be that many branches would not in fact obtain empirical evidence for the correct splitting probabilities (as seen from a global perspective “outside” all the branches), as Adrian Kent pointed out years ago. So, how do we know we’re in a branch where our experiments actually reflect the correct probabilities? Finally, Adam defends these many-worlds views against accusations that they’re unscientific because they’re unverifiable. He properly points out that all scientific theories are unverifiable in the sense of Popper, e.g., deviations in Uranus’s predicted orbit led to the discovery of Neptune, not the overthrow of Newtonian gravity. Later, deviations in the orbit of Mercury did lead to Newtonian gravity being “falsified,” i.e., replaced by a more accurate theory (GR). Here I think Adam’s defense is strained at best. There is a huge difference b/w Newtonian gravity not being falsified by a single apparently discordant measurement (Uranus’s orbit) and the fact that EVERY POSSIBLE measurement outcome is compatible with a theory. To claim the former case is equivalent to the latter is an egregious misrepresentation of the objection of unfalsifiability. To paraphrase one opponent of such views, “Does a theory that predicts everything explain anything?” On to chapter 12!

 

I don't understand, what you think is "not true" in my previous statement. Your nice undergrad-lab experiment described in your paper does not prove quantum nonlocality, or do you claim that its outcome cannot be described by QED? What your experiment indeed demonstrates (as far as I can see from glancing over the paper) are the long-ranged correlations between entangled parts of a single quantum system, which is not contradicting locality of the interactions. QED, as any QT, allows to describe entanglement without violating locality by construction, and also the linked-cluster theorem holds true. It is just careless use of the word "non-locality" instead of "long-ranged correlations" you find very often in the literature, and that is bound to confuse your students rather than helping them to understand that the beautiful Bell-test experiments with photons done in the last 2-3 decades demonstrate that entanglement really means what QT predicts, i.e., the incompatibility of the probabilistic predictions of QT about ensembles with any classical-statistical local deterministic hidden-variable model a la Bell.

 

The formalism for the experimental outcomes is in the paper. That's not the issue. The question is, what is the nature of reality such that those correlations obtain? Simply saying the formalism maps onto the outcomes in no way tells me WHY those correlations obtain, only that you found a formalism that maps onto to them. Again, go to Adam's roulette wheel analogy and the formalism of the paper would equally map to those outcomes. How can that be?

 

QT gives statistical description of entanglement. But the point of Bell theorem is that there is a testable difference between "long-ranged correlations" realized by local physical mechanisms and non-local physical mechanisms when you analyze the data on the event by event basis.
In that sense there is no difference between QT and QED. QED gives its predictions on statistical level and gives no handle for event by event analysis. I suppose that this not so obvious because QFT speaks about "fields" just like electromagnetic field that is considered physical. But the "field" of QFT is not physical. It's statistical.

 

Well, as an experimentalist you should be much less worried about what's reality than the theoreticians, because it's you who defines what reality is! You set up your devices to produce the entangled bi-photon states and the various optical devices and detectors to observe them. What's real is what your detectors show. The theory (in this case QED, simplifying the devices to an effective description which is more or less the same as in classical electrodynamics (quantum optics of optical devices is mostly the hemiclassical approximation, i.e., matter treated phenomenologically in terms of response functions/susceptibilities), except for the detection process of photons itself, which usually is some kind of photoelectric effect (which can be almost always be treated semiclassically, i.e., assuming classical em. fields but quantized electrons). All this is not reality but an (effective) quantum-field theoretical description for the statistical outcome of your detector clicks, and what's real are the clicks, not the theorists' field operators and state operators!

 

QT=Quantum Theory, of which QFT is one realization to describe the electromagnetic interaction in terms of charged particles and the em. field (both quantized quantum fields in the first-principle level of description).

According to QT (and thus also of course QFT) there's nothing else than probabilities. If an observable is not determined through preparation, then it's value is indetermined, and you can only know probabilities for the outcome of measurements of this observable. To test the theory you have to perform experiments on a sufficiently large ensemble to gain enough statistics for the aimed level of statistical significance.

 

You are espousing a variant of the Copenhagen interpretation here. Did you read the book?

 

This is no Copenhagen this is the Minimal Statistical Interpretation, i.e., in my understanding there's no quantum-classical cut (classical theory is a valid approximation to QT due to the sufficiency of coarse-grained observables for macroscopic properties and decoherence) as seems to be the main point of all flavors of the Copenhagen interpretation. As well there's no collapse due to measurement, which is part of some flavors of the Copenhagen Interpretation.

I've not read the book yet. I've to get it first and then (more difficult) also find the time!

 

In Chap 12 Adam does mention retrocausality in passing. He talks about it dynamically, i.e., future outcomes sending information into the past, which is in the early spirit of some adherents, but Aharonov, Price, Wharton, and Cramer have all dismissed this pseudo-time-evolved narrative story at some point (to me personally or in print). As I said earlier, given access to the block universe for explanatory purposes, there's no reason to introduce pseudo-time-evolved explanation, it's superfluous.

 

@RUTA, as you can see from vanhees71's quote, he is indeed not espousing any legitimate variant of Copenhagen, since it has no cut. It is simply not valid quantum mechanics (basically it is a variant of Ballentine's erroneous interpretation).

 

Wow, yes, I totally misread his post #108. He's not claiming there is no Copenhagen interpretation, he's simply claiming HIS interpretation isn't Copenhagen and then explaining why. I'm too tired to read critically today :-) I'll delete my last post. Thnx, atty.

 

I finished the appendix where Adam showed how dBB, MWI, and GRW explain Wheeler’s delayed choice experiment done with an interferometer. The delayed choice was simply to insert the second beam splitter (BS) or not after the photon has passed through the first BS. The explanation is trivially dynamical for these interpretations in this experiment. In order to challenge these dynamical interpretations, you need an experiment like the one shown in Sci Am (below). In that experiment you can choose to insert a lens between photons scattered off electrons passing through a twin slit thereby destroying electron which-way info. If you choose not to insert the lens, the scattered photons carry which-way info on the electron to the photon detector. The electrons make an interference pattern when the lens is inserted and a particle pattern when the lens is not inserted. The lens can be inserted after the electrons have already hit their detector (as in the Kim experiment, below). In the Kim experiment, one could easily say the pilot wave takes info from the first photon (“electron” counterpart) to the second photon (“scattered photon” counterpart) to make sure it goes to the correct detector. But, if a human agent is deciding whether or not to place a lens in front of the scattered photon, as in the Sci Am experiment, then dBB would have to either say the pilot wave is influencing the decisions of the human agent, or that the pilot wave is retrocausal from the lens to the electron.

 

Well, I've not read the book (I've ordered the paper back edition arriving end of May), and it might be unfair against the author to discuss about what's claimed to be in that book in a forum, but if he claims that standard QT implies retrocausality, he's utterly wrong. By the very construction of local microcausal relativistic QFT there cannot be any retrocausality by construction, and so far nothing ever observed hints in this direction!

 

It's your claim that the Minimal Interpretation is errorneous. By repeating this claim, it doesn't become true! The minimal interpretation is all that you need to confront the theory with experiments (at least those realized up to today), and the theory stands all tests. Anything going beyond the minimal interpretation enters the realm of personal world views and thus is not testable by observation and thus is not part of physics but maybe religion. Not that religious believes are unimportant for individuals, but for sure they are not in the realm of science and the part of humane experience described by it.

 

I do not agree, and neither do standard texts like Landau & Lifshitz or Weinberg.

 

Well if you want a causal account of the experiment shown in the Sci Am article, then either the electron hitting the screen causes the agent to insert or not insert the lens (forward causality) or the agent’s decision to insert or not insert the lens causes the electron to hit the screen in the correct place (retrocausality). One might deny that the Sci Am experimental prediction will be seen because a human is making the decision (unlike the Kim et al experiment where beam splitters “make the decision”). QM doesn’t make different predictions based on conscious versus nonconscious intervention so if you believe that, you would be claiming QM (and QFT by extension) is wrong. Hardy proposed an experiment to explore this possibility https://arxiv.org/pdf/1705.04620.pdf

 

How can an interpretation be erroneous?! May be you mean that it is not complete in some sense because it doesn't address some questions?

 

You’ll have to read the book. Adam presents many arguments against physics as whole adopting such an attitude. Once you’ve read his arguments, get back to us as to why you think they’re wrong.

 

It is erroneous because it is self-contradictory. vanhees71 uses a cut, yet he says there is no cut.

Also, he claims to have a solution to the measurement problem that involves neither hidden variables nor MWI, only coarse graining. This is basically a variant of "decoherence solves the measurement problem", which is an error.

And yes, it is an argument from authority - but there is a reason that the standard texts like Landau and Lifshitz or Weinberg use a Copenhagen-like interpretation.

 

I wasn't aware that Ballentine and vanhees71 are the same person!

 

Of course, I don't claim that. My point simply was that all hitherto done experiments with entangled photons and other systems to test Bell's inequality against the prediction of its violation by QT are all fully understood within relativistic local microcausal QFT and thus by construction exclude both spooky-action at a distance and the possibility of retrocausality. All there is is the state preparation in the very beginning which implies the correlations described by entanglement, and all experiments agree with the predictions of QT (particularly relativistic QFT). I don't expect any changes with this conclusion when using humans for the switching decision, but of course one has to do the experiment to be really sure. Physics is indeed an empirical scienc!

 

I don't use a cut. I use real-world macroscopic equipment to prepare states and perform measurement (well, I let my experimental colleagues do that, because I'd for sure mess up the experiment being a theorist ;-)).

I don't claim to solve any "measurement problem". I deny that one exists do begin with for the simple reason that we are able use QT to successfully predict the outcome of measurements (in terms of probability and statistics).

Landau and Lifshitz use indeed a Copenhagen-like flavor, but they hardly discuss interpretational issues at all. Weinberg doesn't take any side but says that the interpretational problem is undecided, although I also fail to see where this apparent problem might be for the reason just given. Weinberg's chapter on interpretation is, however, among the best I've read about the issue (which is as valid for the entire content of this and all his other textbooks). Nevertheless I'm not sharing his opinion on the final dictum on interpretation.

 

That is a cut, because the "macroscopic" equipment is not included in the quantum state.

That alone would be ok (not my position, but certainly one that is coherent and attractive), but you often add that the macroscopic equipment can be included in the quantum state by suitable coarse graining (without hidden variables or MWI) - that would not be ok.

 

Silberstein and I gave a talk at Univ of Maryland on Wed. Afterwards, we had dinner with Jeff Bub and he had some interesting responses to Adam’s book. He was not happy that the book made it seem like he wasn’t aware of Bohm’s interpretation when he was Bohm’s grad student. In fact, Bohm wasn’t taking any more students when Jeff was picking an advisor, but Bohm took Jeff precisely because Jeff had done an undergrad thesis on Bohm’s interpretation. More stories from Bub to follow :-)

 

So did Jeff Bub buy Adam's book?