Cheshire Cat Experiment in July Issue of Nature

[RUTA]

I was wondering what people thought of the Cheshire Cat experiment published in the July 2014 issue of Nature Communications: http://www.nature.com/ncomms/2014/140729/ncomms5492/full/ncomms5492.html. The conclusion is that "any probe system that interacts with the Cheshire Cat system weakly enough will on average be affected as if the neutron and its spin are spatially separated."

 

[Demystifier]

As with any other result obtained by a weak measurement, one should be very careful in interpreting the result as a genuine measurement of a real property of the system.

 

[DrChinese]

RUTA,

I think it makes sense. Quantum sense of course. :)

You can split a variety of quantum things apart and then recombine them to re-construct something. I wrote a paper a while back showing that entangled photons could be split in half (bet you didn't think that was possible) and recombined to restore an entangled state. The quirk is that the final entangled photons were half of Alice and half of Bob (i.e. not the original Alice and Bob). Just as the final objects in the referenced experiment look like the original object.

http://www.drchinese.com/David/EntangledFrankensteinPhotonsA.pdf

See figure 5.

 

[DrChinese]

By the way, do you have any new papers coming along? Haven't seen anything new lately... :) :) :)

 

[RUTA]

The reason I asked is because I can compute the intensities at O and H when MAG is introduced on path I or II (as given by their Eqs 13-16) in a local fashion. That is, all contributions to the amplitudes at O and H are computed locally to the neutrons and their spins at all times in the experiment. All I did to obtain the intensities they give was compute the amplitudes at O and H then square appropriately (only |-> at O, both |+> and |-> at H). To compute the amplitudes, I apply the unitary operator given in their Eq 8 at the appropriate place on I or II, add the chi phase factor to each path as shown on page 4, add reductions for beam splitters, and add a factor of i for reflections off mirrors or beam splitters through the interferometer, i.e., standard interferometer calculation. When I compute the amplitudes in this fashion, it's easy to see that the reason for the difference in the intensity at O for MAG added to path I or II comes from the fact that the spin z Pauli matrix turns a |+> spin x eigenstate into a |-> spin x eigenstate and vice-versa. So, when you add MAG to the path with |-> you get |-> with a |+> contribution, which doesn't change the |-> intensity at O. But, when you add MAG to the path with |+> you get |+> with a |-> contribution, which does change the |-> intensity at O. This is the only mystery for the experiment, since you trivially get a reduction in the |-> count at O when you block the path with |->. The bottom line is, my computation of the MAG intensities in no way implies the separation of spin from its host neutron through the device at any time. Thus, I have no idea why they’re making that claim.

 

[RUTA]

By the way, do you have any new papers coming along? Haven't seen anything new lately... :) :) :)

We revised an arXiv paper on 30 Aug 14 explaining our approach to quantum gravity and unification. That paper is scheduled to appear in a book on quantum spacetime in shorter form (http://arxiv.org/abs/0908.4348) sometime next year. My philosophy colleague has just begun work on his part of a paper we will submit to Quantum Studies based on my calculations and RBW explanations of twin-slit interference, a simple EPR-Bell set up, Dowker's GHZ set up (), and the DFBV experiment (). I'll email you a copy of that when it's done. Thanks for asking :-)

 

[RUTA]

The reason I asked is because I can compute the intensities at O and H when MAG is introduced on path I or II (as given by their Eqs 13-16) in a local fashion. That is, all contributions to the amplitudes at O and H are computed locally to the neutrons and their spins at all times in the experiment. All I did to obtain the intensities they give was compute the amplitudes at O and H then square appropriately (only |-> at O, both |+> and |-> at H). To compute the amplitudes, I apply the unitary operator given in their Eq 8 at the appropriate place on I or II, add the chi phase factor to each path as shown on page 4, add reductions for beam splitters, and add a factor of i for reflections off mirrors or beam splitters through the interferometer, i.e., standard interferometer calculation. When I compute the amplitudes in this fashion, it's easy to see that the reason for the difference in the intensity at O for MAG added to path I or II comes from the fact that the spin z Pauli matrix turns a |+> spin x eigenstate into a |-> spin x eigenstate and vice-versa. So, when you add MAG to the path with |-> you get |-> with a |+> contribution, which doesn't change the |-> intensity at O. But, when you add MAG to the path with |+> you get |+> with a |-> contribution, which does change the |-> intensity at O. This is the only mystery for the experiment, since you trivially get a reduction in the |-> count at O when you block the path with |->. The bottom line is, my computation of the MAG intensities in no way implies the separation of spin from its host neutron through the device at any time. Thus, I have no idea why they’re making that claim.

In http://arxiv.org/pdf/1409.0808.pdf the authors write:

When a magnetic field is applied in path I changing the neutron spin state from |+> to a|-> + b|+>, the |-> component of the wavefunction of this
can interfere with the wavefunction of path II, such that the counts in both
detectors should depend on chi. When a magnetic field is applied in path II changing the neutron spin state |-> to c|-> + d|+>, the |+> component
of the wavefunction of this path can interfere with the wavefunction of path I, such that the counts in D2 should depend on chi. But since the detector D1 selects only the |-> component of spin, the counts in this detector should not depend on chi. All these predictions are confirmed by the experiments.

The behaviors described above led the authors to say:

"(...) an absorber with high transmissivity has on average no significant effect on the measurement outcome if it is placed in path I. It is only effective if it is placed in path II. In contrast to that, a small magnetic field has on average a significant effect only in path I, while it has
none in path II. Therefore, any probe system that interacts with the Cheshire Cat system weakly enough will on average be affected as if the neutron and its spin are spatially separated" [14].

This is what they mean when say that "the experimental results suggest that the system behaves as if the neutrons go through one beam path, while their magnetic moment travels along the other" [14]. But as we have seen here, the results can be explained as simple quantum interference, with no separation between the neutron and its spin.

 

[RUTA]

Here is my analysis:

http://arxiv.org/abs/1410.1522

It was just posted on the arXiv last night, thus the delay in my sharing it. I submitted it to Nature Communications where it's "under consideration," but it's "safe." It's simple quantum mechanics (which I've taught many times), but I had a couple colleagues check and sign off on it just to be sure.

 

[RUTA]

Much has happened concerning the topic of quantum Cheshire Cat (qCC) in the past year. If you don't care about that story, you can just read the final version of the paper that was recently accepted for publication in International Journal of Quantum Foundations http://www.ijqf.org/archives/3013. Here is the story behind this paper.

Nature Comm said our Brief Communication Arising (restricted to 600 words) presented a new concern for weak values that needed to be established in a peer-reviewed journal on its own merit. Specifically, the weak values per se do not suffice to establish the qCC interpretation, those weak values must be measured in the context of a linear interaction in order to reasonably infer qCC. So, we submitted a regular paper (in various forms) to American Journal of Physics, Foundations of Physics, and Physical Review A, in turn. AJP and FoP sent the paper to 4 referees, 3 recommended publication, 1 said the paper was correct, but too contentious for undergrad physics majors (AJP). All three editors said the paper was a Comment and needed to be published in Nature Comm. We resubmitted the BCA to Nature Comm with all the referee reports and editorial replies. Nature Comm said, you guessed it, the paper is making a new claim that needs to be established independently. So, we submitted a paper to New Journal of Physics where Aharonov's original qCC paper and Correa et al.'s Comment thereupon appeared. They said, you guessed it, the paper is a Comment and needs to be published in Nature Comm. We appealed and they stuck by their claim.

I ran into Vaidman and Elitzur at a conference in Vaxjo in June and discussed the Denkmayr et al. experiment. I was going to present an update on Relational Blockworld, but quickly realized there was much more interest in weak values in general and the qCC experiment in particular, so the conference organizer (Andrei Khrennikov) let me change my presentation to what is essentially contained in the paper linked above. One of the experimentalists in Denkmayr et al. (Sponar) presented in my session immediately after me and he confirmed that my understanding of their experiment was accurate. After the conference, I realized there was some confusion among those involved as to what was actually required to establish the qCC interpretation, so I concluded that Nature Comm's assessment was correct, the technical point needed to be vetted in and of itself.

Thus, we wrote the paper linked above and submitted it to IJQF (where Vaidman and Aharonov are members), sharing this history with the editor (Shan Gao). The editor agreed that indeed the paper isn't a Comment, but is arguing for a technical point about the interpretation of weak values in the qCC experiment. After revising the paper per their comments, the two referees recommended publication and the editor agreed, so the paper is with copy editing now.

The bottom line is, as I state in my IJQF post linked above: We believe the weak interaction of the quantum Cheshire Cat experiment must be linear if the weak values are to support the quantum Cheshire Cat interpretation, i.e., that a particle and one of its properties are spatially separated. For example, even though Denkmayr et al. measured the proper weak values for quantum Cheshire Cat in their neutron interferometry experiment, they did so using a quadratic interaction. We show how the quadratic interaction in their experiment destroys the quantum Cheshire Cat interpretation. Indeed, if the quantum Cheshire Cat interpretation is established per the weak values alone, regardless of the degree of interaction, then Denkmayr et al.'s experiment would be a reductio against the interpretation.