## Thursday, October 13, 2011 ... //

### Physics Today: quantum mechanical wars

A new issue of Physics Today (which is affiliated with the American Physical Society) offers some texts about black hole mergers, lab safety standards for students, and all kinds of things such as nanoscale electrochemistry, nanowires, analyses of kidney diseases, and dozens of other things.

However, as you may expect, I will pick two topics, especially the second one. ;-)

Two of the texts are promoting the global warming propaganda in new, unusual ways, and I admit that I haven't read the whole texts and I don't even recommend you to do so unless your nerves may really afford it.

A Steven Sherwood claims that the global warming fearmongers are revolutionaries just like Galileo or Copernicus and the skeptical scientific community and the skeptical public just isn't able to appreciate their revolutionary findings.

Just imagine how incredible such a statement is in combination with the proclamations we used to hear just a few years ago when it was fashionable to parrot nonsense about a looming climate cataclysm. We would be hearing that the main argument for the climate alarmism is nothing else than the scientific consensus: the climate scientists were as certain and as unified as the Catholic Church defending the geocentric system.

We would hear that there may sometimes exist mavericks and naysayers such as Galileo but in 97%-98% of the cases, the scientific consensus is right. And of course, everyone is obliged to join the scientific consensus and become a fearmonger because it was the only politically correct attitude.

Quite suddenly, when people are learning that no one except for ideologically blinded bigots and hired guns believes the climate catastrophic rubbish, the self-evidently small group of climate alarmists promotes mavericks and minorities. You can't have it both ways. Moreover, none of the two arguments is valid. Whether someone is in a majority or minority isn't a scientific argument for or against a scientific hypothesis. Such arguments don't work in either way.

Richard Somerville (one of the alarmist losers who once lost a debate against the skeptics) and Susan Hassol wrote an article that is not free. The abstract says that the climate alarmists are obliged to learn every lesson that Goebbels may teach them in brainwashing the public. Very nice.

Quantum wars

But aside from dimwits, there is some smarter yet confrontational stuff. Note that Physics Today doesn't allow doubts about a coming climate catastrophe – something that only complete idiots are ready to believe – but it legitimizes discussions about the foundations of quantum mechanics.

There have been lots of Physics Today articles about the foundations of quantum mechanics in the recent 12 months. The most recent controversies build upon these articles:

Original irritating article
A time‐symmetric formulation of quantum mechanics by Yakir Aharonov, Sandu Popescu, and Jeff Tollaksen (abstract): November 2010

Criticisms of the text above:
Time-symmetric quantum mechanics questioned and defended by Art Hobson
Time-symmetric quantum mechanics questioned and defended by Michael Nauenberg

Defense of the first text, criticism of criticism
Consistent treatments of quantum mechanics by David Mermin
Consistent treatments of quantum mechanics by Yakir Aharonov, Sandu Popescu, and Jeff Tollaksen: they thank Mermin

A defense of Consistent Histories against the first text
Consistent treatments of quantum mechanics by Robert Griffiths
You may see that the two armies are pretty much equally large. ;-) I view myself as a champion of the Consistent Histories although the difference between the approaches to quantum mechanics that I could endorse is often just about the emphasis and marketing.

So it shouldn't be surprising that I agree with Griffiths and consider anti-Consistent-Histories remarks by Aharonov et al. unjustifiable. Griffiths recalls that Consistent Histories solve various measurement "paradoxes" and also offer you a consistent scheme to ask questions about probabilities what happened throughout the time, including what happened before the measurement.

While I have often endorsed David Mermin's views on quantum mechanics, I am much closer to the criticisms written by Nauenberg and by Hobson. They localize statements by Aharonov et al. that really seem like one of those typical misunderstandings of quantum mechanics. Nauenberg of Santa Cruz (I had to meet him in H1 of 2000 which I spent in UCSC) shows that Aharonov et al. seem to think that one may completely determine $S_x$ and $S_z$ at the same moment because one may measure both observables (one by one). Of course, this is bullshit: the first measurement changes the state. Also, Nauenberg is right that one can't divide ensembles to arbitrary "classical subsets" – exactly because the observables don't commute and the relative phases are important.

Hobson of Arkansas also criticizes Aharonov et al. for underestimating the role of relative phases (yes, that's quite ironic for Aharonov to ignore phases haha). Hobson suggests that they even seem to believe that a particular combination of states "up" and "down" may be summarized as 50% odds for one option and 50% odds for another option. However, that's not the case: the physical properties of the system depend on the relative phase as much as they depend on the ratio of absolute values (the "odds").

Mermin would agree with Hobson and Nauenberg (and their points should be addressed, Mermin says) except that he also thinks that these two critics misunderstood the article by Aharonov et al. who don't make these errors (and who were just incomprehensible). I am not sure. The flawed sentences that Hobson and Nauenberg criticize seem pretty clear and clearly wrong to me.

The sociology of the foundations of quantum mechanics finds itself in a messy state. They really analyze – and yes, constantly try to undermine – something that's been understood by the good physicists since the mid and late 1920s. Some people, like Mermin, added (or refined) pedagogical and refreshing Gedanken experiments that the fathers of quantum mechanics would surely like as well but I think that they could also correctly answer them.

Some other people, like Griffiths or Zurek (but also Gell-Mann, Hartle, Omnes, and others), added insights that clarify the emergence of "classical intuition" from quantum mechanics for large or otherwise "nearly classical" objects and that allow us to ask and answers "more general questions" than questions about separate measurements (like the probabilities of histories, and it's explained when this is possible). Those things make sense. But I think that the rest just a sequence of deep misunderstands of quantum mechanics and constant attempts to reshape it in the old picture of classical physics. I think it's the case of Aharonov et al., too.

Aharonov et al. emphasize the notion of the "weak measurement". It's the whole point to talk about it in this experiment, they say. It's interesting because Mermin who defends them doesn't mention "weak measurements" at all. ;-) I think that this whole concept of a "weak measurement" is just hot air. It's another piece of propaganda that is meant to allow the confused people to pump classical reasoning into quantum mechanics.

The "weak" adjective is there to allow them to think that things are "still quantum mechanical" while the "measurement" is there so that they can imagine that things follow the classical intuition throughout the evolution. But they don't. One may introduce weak interactions with the environment or a measuring apparatus which will weaken the interference and accelerate the emergence of a "classical description". But there's still the quantum behavior in such systems and quantum mechanics allows us to calculate what happens (as long as we formulate the questions accurately, i.e. convert them to a well-defined measuring apparatus), anyway.

The "difficult experiment" that Aharonov et al. discuss is simply a measurement of spin, first $S_x$, then $S_x$ or $S_z$ (we don't know which), and then again $S_x$. Well, if we measure $S_x$ thrice, we get the same result thrice. If we measure $S_z$ in between, the first and third measurements are uncorrelated. OK and what?

Mermin tries to make the statements by Aharonov et al. less controversial but I actually disagree with Mermin's versions of the statements as well.
Mermin: For example, Aharonov and coauthors say, “The results at $t$ depend... on what happens later at $t_1$.” Nauenberg says, and it sounds right, that this contradicts standard quantum mechanics. He surely would not have objected if the authors had instead said, “What we can learn about the results at $t$ depends... on what we learn later at $t_1$.” But the way they do state it suggests causality acting backwards in time, which would indeed contradict quantum orthodoxy.
They don't just "suggest" a violation of causality: they boldly and unambiguously state it. And Mermin is actually doing the same thing. His softening simply didn't help. What we can learn at an earlier time $t$ can't be affected by what we can learn at a later time $t_1$. For example, we can't predict the winner of the Republican primaries now whether or not we know whether Iran will destroy America by nuclear weapons before the 2012 presidential elections – especially because we don't know the latter (the distant future that will depend on the less distant future). Of course, at a later time $t_1$, we may give richer interpretations of what happened at the earlier time $t$ and how it is correlated with things that happened at $t_1$. But if you're interested about events and properties at $t$ only, events at the later time $t_1$ can't change it at all. This is what causality means and quantum mechanics rigorously obeys it!

Wikipedia article on Aharonov presents Aharonov's views on causality very clearly and unambiguously (and, I think, fairly):
In 1988 Aharonov et al. published their theory of weak measurement, which doesn't disturb the quantum state being observed. This work was motivated by Aharonov's long time quest to experimentally verify his theory that apparently random events in quantum mechanics are caused by events in the future. Verifying a present effect of a future cause requires a measurement, which would ordinarily destroy coherence and ruin the experiment. He and his colleagues were able to make weak measurements and verify the present effect of the future cause.
Some badly behaved coherence kills Aharonov's theory on time-machine dictatorships but he surely thinks he can resuscitate it by making measurements-non-measurements. ;-) No, you can't. It's just rubbish.

But let's return to the XXX and XZX experiment.

In this experiment, every measurement completely reshapes the wave function to a pure known spin eigenstate. If the next measurement measures the spin with respect to the same axis, we get the same result. If it measures the spin with respect to an orthogonal axis to the previous axis, we get 50%-50% odds. The predictions always depend on the previous axis only; they never depend on the future axis.

In the same way, I think that Aharonov et al. – and implicitly Mermin – are just wrong when they say that a measurement with respect to the diagonal axis "in between $x$ and $z$" is effectively a measurement of both $S_x$ and $S_z$. It's not. The diagonal measurement is sensitive on the relative phases of the up-and-down components in the $S_x$ eigenstate basis; and it's also sensitive on the relative phases in the $S_z$ eigenstate basis. The measurement of one diagonal component is just one measurement, not two. Also, it brings the particle into an eigenstate of $S_x+S_z$ which are surely neither eigenstates of $S_x$ nor eigenstates of $S_z$.

These are just examples of technical and conceptual errors that completely destroy a paper that builds upon them. I don't understand how someone may pretend that they don't matter. I can't get rid of the feeling that these (wrong) technical comments are there just as fog that is supposed to legitimize some big (and wrong) philosophical claims that can't really be defended in any rational way. Aharonov et al. surely want to make ambitious claims – e.g. that time may go in both directions in quantum mechanics, whatever it is supposed to mean – but the actual detailed evidence underlying these big claims is completely wrong. I don't understand what Mermin thinks that Aharonov et al. discovered; I don't understand why he's trying to defend them. In other contexts, talks, and papers, Mermin would correctly explain why people like Aharonov are just wrong. What does he "see" that they discovered about the XXX or XZX measurement? With his pedagogical mastery, Mermin could surely explain it if a discovery existed but he hasn't done so: at least I see nothing of value in the article by Aharonov et al.

In some sense, I think that the "big claim" they want to prove by the flawed detailed science is analogous if not equivalent to the crackpottery that Sean Carroll and other people misunderstanding thermodynamics have been promoting for years. They also seem to misunderstand the concept of the logical arrow of time: one may prepare a physical system in the initial state and consistently predict well-defined and unique probabilities that it will have a property in the final state. Quantum mechanics calculates such predicted probabilities as squared absolute values of the complex amplitudes.

However, one cannot prepare a state in the final state and calculate the probabilities that it had some properties in the initial state. We never "possess" the future. We can't grab it. Our ability to "select" the initial state is always greater than our ability to "select" the final state. (Is this reverted logic what they call postselection?) To say something about the initial state (probabilities of some of its properties), one needs to make a "retrodiction" which is something completely different than a prediction. In particular, it is a form of Bayesian inference in which the results depend on arbitrary prior probabilities that always have to be incorporated and that always bring some subjectivity to the calculation. Logic doesn't have a symmetry between assumptions and their consequences – the two propositions linked in an implication. For the same reason, the logic of the Universe doesn't have a past-future symmetry even if the microscopic laws are T- or CPT-symmetric. The logical arrow of time guarantees that such a would-be symmetry is completely invalid.

That's just a different way of saying that the phenomena of the real world – where information is being lost and forgotten, which is true especially in the macroscopic world – are self-evidently irreversible. If you find a cold soup on the table, you can't retrodict when it was warm (by looking at the soup). Although Aharonov et al. use a different language and come from a different culture, I am pretty much convinced that they misunderstand this elementary physics (the origin of the second law of thermodynamics and related things) much like Carroll does (even though they don't focus on thermodynamics: they attack the existence of the logical arrow of time directly). What about Mermin? I don't know. In this battle, I am much more convinced that Nauenberg and Hobson – and Griffiths - are those who know what they're talking about and Mermin is a man who grew tolerant to complete nonsense.

And that's the memo.

#### snail feedback (5) :

One doesn't need to be a physicist to understand that the future cannot affect the present, but anticipation of the future can. It is waht allows an outfielder to catch a baseball.

So, what Harlow did 'de facto say' was that past futures can predict future futures. :)

Wasn't it a Sherwood that showed up in Svensmark's youtube who, following Henrik Svensmark's presentation declared to Svensmark that it was "irresponsible" ? (No science issue brought up there either.)