Can Penrose's Felix experiment prove the role of gravity in objective collapse?

[vanesch]

As I was reading in "The road to reality" (Penrose), where he describes his Felix experiment, to find out whether gravity gives rise to objective collapse or not, it occurred to me that he - IMHO - is overlooking something.

FYI, the Felix experiment (in different variations) goes as follows:
a photon is directed towards a beamsplitter, and one of the beams will go to a rigid, light mirror before going into a long waveguide, while the other beam directly goes into a waveguide. At the end of both waveguides, there's a hard mirror, bouncing back the beams. The light mirror is suspended elastically in such a way, that it has a mechanical period equal to the double traveling time in the waveguide: as such, it is in exactly the same position when the beam comes back onto it. The mirror is set into oscillatory motion by the momentum of the photon when the photon takes this road, while it is not set in motion when the photon doesn't take the road, but, as said, the suspension of the mirror is such that it comes back to the original position at the moment when the beam bounced back on it (restoring the exact original momentum).

Penrose reasons, that if the photon is still in a superposition of the two "arms" after bouncing back, it will interfere again on the beamsplitter, and as such, only ONE outcome will be possible, namely that the photon "recombines" and goes back into the source. A detector on the other sending side of the beamsplitter will then never register a click.
His argument is that during the flight time of the beam, the mechanical position of the light mirror is in a superposition of 2 positions, namely "oscillating" and "not oscillating" mechanically (because the photon is in a superposition of branch 1 and branch 2 of the guides).
If, Penrose argues, the moving mass of the mirror gives rise to objective collapse (his OR scheme) because of the two different gravitational configurations, then the photon will also collapse into one of the two states, and will not interfere with itself anymore at the beamsplitter upon return from the two beams. And as such, in 50% of the cases, the photon will return into the source, and in 50% of the cases, it will go into the detector.

But after pondering a bit about it - and leaving all technicalities aside (Penrose clearly isn't an experimentalist !) - I think he forgot environmental decoherence. When the mirror is oscillating mechanically (half a period), chances are great that the mirror will in the mean time interact with the environment (a gas atom, background radiation, whatever), and this interaction will of course be different from the interaction when the mirror is not oscillating. So I think that IN ANY CASE no interference will be seen, because of the different environmental states which get entangled with the mirror position ; as such, the lack of interference is, to me, not a proof of objective collapse.

Any comments ?

cheers,
Patrick

PS: I'm not often on the net for the moment: I'm traveling and have not much possibility to get a link...

 

[Careful]

But after pondering a bit about it - and leaving all technicalities aside (Penrose clearly isn't an experimentalist !) - I think he forgot environmental decoherence. When the mirror is oscillating mechanically (half a period), chances are great that the mirror will in the mean time interact with the environment (a gas atom, background radiation, whatever), and this interaction will of course be different from the interaction when the mirror is not oscillating. So I think that IN ANY CASE no interference will be seen, because of the different environmental states which get entangled with the mirror position ; as such, the lack of interference is, to me, not a proof of objective collapse.
Any comments ?
cheers,
Patrick
PS: I'm not often on the net for the moment: I'm traveling and have not much possibility to get a link...

The experiment is supposed to be conducted in circumstances where there is no interaction with the evironment, moreover there is no guarantee that a realtistic EVD model will do that job for you unless you propose detailed interactions. The problem with EVD is that, in principle, you can ``explain´´ even two contradictory things as long as you put in the ``correct´´ interaction Hamiltonians. But I shall return the question to you, if no distiction with QM is to be found experimentally, why not use Penrose OR theory then since it is clearly superior in my opinion ??

 

[caribou]

Hehehe... I just Googled "Penrose objective collapse" to read up a little before replying and the top search result was this thread!

But after pondering a bit about it - and leaving all technicalities aside (Penrose clearly isn't an experimentalist !) - I think he forgot environmental decoherence. When the mirror is oscillating mechanically (half a period), chances are great that the mirror will in the mean time interact with the environment (a gas atom, background radiation, whatever), and this interaction will of course be different from the interaction when the mirror is not oscillating. So I think that IN ANY CASE no interference will be seen, because of the different environmental states which get entangled with the mirror position ; as such, the lack of interference is, to me, not a proof of objective collapse.

Yeah, I'm just thinking again of the famous example of the 1 millimeter-wide dust grain deep in interstellar space decohering from a superposition of two positions 1 millimeter apart due interactions with the background radiation from the Big Bang alone in only about a billionth of a second.

To get into a superposition of two positions before decoherence in the time the mirror has in a less isolated place like a laboratory, I guess for the experiment to work that this has got to be a very small and fast-moving mirror indeed.

 

[vanesch]

But I shall return the question to you, if no distiction with QM is to be found experimentally, why not use Penrose OR theory then since it is clearly superior in my opinion ??

Well, as far as I know, it is still inexistant !

 

[CarlB]

Here's the paper proposing the experiment:

Towards quantum superpositions of a mirror
William Marshall, Christoph Simon, Roger Penrose, Dik Bouwmeester
http://www.arxiv.org/abs/quant-ph/0210001

They think it can be done.

My guess is that if you can solve the construction problems, the result will, in fact, be an apparent superposition. I say "apparent" because I (unlike most physicists) believe that wave function collapse preserves quantum phase information and consequently is not detectable. In other words, quantum mechanics will win once again (and in this I agree with most physicists but for a different reason). I found Penrose's arguments about about the interaction between gravitation and wave function collapse to be very persuasive, but if you assume that wave function collapse preserves phase, it becomes undetectable by a measurement anyway.

Carl

 

[Hans de Vries]

Superposition bookkeeping catastrophe.

.
.
There seems to be a bookkeeping catastrophe looming here for
certain classes of QM interpretations.Imagine a 10 mW laser sending out 2.5 10¹⁶ photons per second.
The first photon passes and leaves the system in a superposition of
two states, which only becomes resolved when the photon is detected.
The second photon passes which doubles the number of superposition states.
Each next photon again doubles the number of superposition states...

Imagine this going on with the above rate of photons passing...

A passing photon actually would leave each point of the vacuum it passes
in a doubled superposition of states. (Pair creation is field dependent)
Interstellar photons can take billions of years to become detected,
if ever... Regards, Hans

 

[DrChinese]

Question: I assume that if the experiment were run and the results accepted, and the result was that there WAS decoherence as a result of Penrose's OR: Then that would be evidence that gravity is a quantum force. On the other hand, if there were no OR then gravity is not a quantum force. Is that the idea?

 

[vanesch]

My guess is that if you can solve the construction problems, the result will, in fact, be an apparent superposition.

The funny thing is that if you treat the entire system totally classically, you ALSO get interference. Only, we now have HALF of the momentum on the mirror (swing will be half of the quantum version in the branch where we have impact on the mirror), but this doesn't change the classical oscillation period.
So I see the following situations:
*) interference observed:
- quantum superposition remains valid beyond the OR scheme ;
or:
- we have a purely classical light beam with classical Poynting momentum transfer
*) no interference observed:
- OR scheme valid
or:
- we have some uncontrolled loss of phase coherence (experimental difficulty)
or:
- we have entanglement with some environment (including classical gravity!) - the classical equivalent of which is of course some uncontrolled loss of phase coherence
So it is hard to say 1) what will be the eventual outcome (of course, the most probable outcome of the first trials is no interference, due to uncontrolled loss of phase coherence) and 2) given that outcome, what to conclude from it.
Mind you, I would also hope for a kind of gravitationally induced objective reduction, but my impression is that this kind of experiment is "too cheap" to establish violation of unitarity.

 

[Spin_Network]

As I was reading in "The road to reality" (Penrose), where he describes his Felix experiment, to find out whether gravity gives rise to objective collapse or not, it occurred to me that he - IMHO - is overlooking something.
FYI, the Felix experiment (in different variations) goes as follows:
a photon is directed towards a beamsplitter, and one of the beams will go to a rigid, light mirror before going into a long waveguide, while the other beam directly goes into a waveguide. At the end of both waveguides, there's a hard mirror, bouncing back the beams. The light mirror is suspended elastically in such a way, that it has a mechanical period equal to the double traveling time in the waveguide: as such, it is in exactly the same position when the beam comes back onto it. The mirror is set into oscillatory motion by the momentum of the photon when the photon takes this road, while it is not set in motion when the photon doesn't take the road, but, as said, the suspension of the mirror is such that it comes back to the original position at the moment when the beam bounced back on it (restoring the exact original momentum).
Penrose reasons, that if the photon is still in a superposition of the two "arms" after bouncing back, it will interfere again on the beamsplitter, and as such, only ONE outcome will be possible, namely that the photon "recombines" and goes back into the source. A detector on the other sending side of the beamsplitter will then never register a click.
His argument is that during the flight time of the beam, the mechanical position of the light mirror is in a superposition of 2 positions, namely "oscillating" and "not oscillating" mechanically (because the photon is in a superposition of branch 1 and branch 2 of the guides).
If, Penrose argues, the moving mass of the mirror gives rise to objective collapse (his OR scheme) because of the two different gravitational configurations, then the photon will also collapse into one of the two states, and will not interfere with itself anymore at the beamsplitter upon return from the two beams. And as such, in 50% of the cases, the photon will return into the source, and in 50% of the cases, it will go into the detector.
But after pondering a bit about it - and leaving all technicalities aside (Penrose clearly isn't an experimentalist !) - I think he forgot environmental decoherence. When the mirror is oscillating mechanically (half a period), chances are great that the mirror will in the mean time interact with the environment (a gas atom, background radiation, whatever), and this interaction will of course be different from the interaction when the mirror is not oscillating. So I think that IN ANY CASE no interference will be seen, because of the different environmental states which get entangled with the mirror position ; as such, the lack of interference is, to me, not a proof of objective collapse.
Any comments ?
cheers,
Patrick
PS: I'm not often on the net for the moment: I'm traveling and have not much possibility to get a link...

One can see Penrose explain in detail here:http://streamer.perimeterinstitute.ca:81/mediasite/viewer/

there are some interesting things Penrose devulges in his recent book.

 

[George Jones]

I find Penrose's views on this stuff to be tremendously interesting.

Various version of seminar presentations by Penrose are available on the web . I'm having trouble getting into the site given by Spin_Network, so here's http://cgpg.gravity.psu.edu/online/Html/Seminars/Fall1998/Penrose/Slides/s01.html" (maybe a bit old). There is at one link that has audio - I'll try and find it.

The original reference for his ideas about gravity's role is

R. Penrose, "On Gravity's Role in Quantum State Reduction," General Relativity and Gravitation, 28(5), 1996, pp. 581-600.

This is reprinted as Chapter 13 of Physics Meets Philosophy at the Planck Scale.

I hope the experiment outlined in Carl's link gives unequivocal results, but I fear that the results, due to the points given by vanesch, may not by be clear cut. If the experiment gives clear cut results, then my views are the same as Vanesch - I hope for verification of Penrose's ideas, but I doubt this will be the case.

We need (just for excitement's sake) an experimental result that rocks physics to its core. Maybe Higgs won't turn up at the LHC.

Regards,
George