# Experiment to test if many worlds is the correct interpretation

1. Jul 18, 2013

### jezemine

hi there. I studied physics for years, even went as far as getting a PhD. but that was 13 years ago, since then rusty. Anyway something occurred to me and I wondered what you all would think about it.

My question is, can anyone think of an experiment you could do to check if the "many worlds" interpretation (the one I like best) is the "correct" one?

For example you might imagine this: let's say the universe splits into two, but somehow the two "paths" manage to come back together so that the two "worlds" can interfere with one another. Same idea as a single photon interfering with itself in a double-slit experiment, except instead of a photon interfering with itself, it's the entire universe.

I was thinking maybe there's some experiment you could do in one universe that could detect something akin to the interference fringes of entire universes interfering with one another. Is this crazy?

Perhaps this can't happen due to decoherence? it's hard to get much more macroscopic than the whole universe I suppose. Decoherence is always killing inteference patterns. OTOH decoherence comes from degrees of freedom that you aren't following right, ones that are connected to the environment. if the system is the entire universe then there is no "environment" to get mixed up with.

One thing's for sure: Hilbert space is big! there aren't even words for how big it is!

Last edited: Jul 18, 2013
2. Jul 19, 2013

### Demystifier

It's possible in principle, but as you said, decoherence makes it impossible in practice.

3. Jul 19, 2013

### kith

I don't think this would test the MWI. If we have coherence again, the interpretation of the interfering states as different universes is not meaningful anymore.

I suspect that any interpretation which allows a splitting of worlds, must also allow the recombination of worlds. There are experiments where a cyclic behaviour wrt to the coherence of the system occurs; Rabi oscillations for example. These are destroyed by spontaneous emission but under certain experimental conditions, there are revivals.

4. Jul 19, 2013

### audioloop

if is confirmed a theory of objective reduction (collapse model) is indirectly refuted.

.

5. Jul 19, 2013

### Vanadium 50

Staff Emeritus
If MWI is truly an interpretation, there is, by definition, no experiment can distinguish them. That would require MWI to be a theory in competition with ordinary QM.

6. Jul 20, 2013

### tom.stoer

My interpretation of the MWI was always that all interferences are direct consequences of the MW. The different branches a just the different components of the wave function. So there is nothing beyond these interferences and therefore I agree with Vanadium 50.

Am I wrong?

7. Jul 20, 2013

### Staff: Mentor

What is "ordinary QM" (or what is not ordinary in MWI)?

MWI proposes that the evolution of the wavefunction is everything that exists - in other words, that you can apply QM to everything. There is no possible test which could violate QM and be in agreement with its many-worlds interpretation.

8. Jul 20, 2013

### audioloop

no so fast...

"Imagine....
every branch can split to a Graham's number exponential of branchs, in turn, any split of that branch split again to a Graham exponential and so on and so on.......
per secula seculorum......."

https://www.physicsforums.com/showpost.php?p=4438383&postcount=14

9. Jul 21, 2013

### Gerinski

David Deutsch argues that the double-slit experiment is 'the experiment which proves MWI' since in his point of view MWI is the only tenable explanation for what we observe in the experiment.

"According to the many worlds interpretation, each particle interferes with another particle going through the other slit. What other particle? Another particle in a neighbouring universe. In my opinion, the argument for the many worlds was won with the double-slit experiment. It reveals interference between neighbouring universes, the root of all quantum phenomena."

and quantum computation

"One day, a quantum computer will be built which does more simultaneous calculations than there are particles in the Universe. Since the Universe as we see it lacks the computational resources to do the calculations, where are they being done? It can only be in other universes. Quantum computers share information with huge numbers of versions of themselves throughout the multiverse. Quantum computers are the first machines humans have ever built to exploit the multiverse directly. At the moment, even the biggest quantum computers can only work their magic on about 6 bits of information, which means they exploit copies of themselves in 26 universes-that's just 64 of them. Because the computational feats of such computers are puny, people can choose to ignore the multiverse. But something will happen when the number of parallel calculations becomes very large. If the number is 64, people can shut their eyes but if it's 1064, they will no longer be able to pretend."

10. Jul 21, 2013

### tom.stoer

I think Deutsch's interpretation is consistent.

But the same experiment can be explained with some collapse-interpretation; that's why an interpretation is an interpretation; it's a matter of taste, philosophy, and perhaps Ockham's razor ...

Last edited: Jul 21, 2013
11. Jul 21, 2013

### stevendaryl

Staff Emeritus
Well, I think people are a little fuzzy about what's "just an interpretation" and what's a different theory. What I consider "standard quantum mechanics" has the following components:

1. You choose either a mixed state or a pure state to represent the initial state of the system of interest. You normalize this state.

2. Evolve the state forward in time using the Schrodinger equation (or something similar) to g

3. Perform a measurement M, which corresponds to some Hermitian operator O.

4. The result will be $r_i$ with probability $P_i$, where $P_i$ is the norm of the projection of the state onto the subspace having eigenvalue $r_i$.

5. After the measurement, the state of the system is described by the projected state (renormalized).

I think that variants of quantum mechanics (MWI, Bohm, stochastic models, etc.) differ in what happens at step 5. Does measurement collapse the wave function? When and how does that happen? They answer that question in different ways. So they are actually different theories, not just different interpretations of the same theory. To get experimental data about which theory is correct or incorrect would involve testing whether the wave has collapsed in step 5. And the only way to test that is to look for interference effects (there are none if the wave function has collapsed).

Unfortunately, it is practically impossible to detect interference effects involving macroscopic objects like detectors (or scientists, or dead cats). So even though the various variants of quantum mechanics are different theories, not just different interpretations, there is no practical way to experimentally distinguish them. So for all practical purposes, they are different interpretations of the same theory, since they make the same (testable) predictions.

12. Jul 21, 2013

### Delta Kilo

How about a proper Shrodinger cat experiment? The idea is to see collapse from the p.o.w of the cat inside the box and superposition from the p.o.w of the experimenter outside of the box.

Consider the following experimental setup:
* Polarizing beam splitter sends incoming 'test' photons into detectors D+ or D-
* Classical memory element M is set to 0 when D- is triggered and to 1 when D+ is triggered
* Gate G controls polarization of the 'readout' photons: leaves it unchanged when M=0 and rotates polarization by 90 degrees when M=1. (Assume there is no relative phase shift introduced by the gate G, otherwise it can be measured and compensated for elsewhere)

Let's try this:
* Send a few vertically polarized readout photons, measure the result at 0 and 45 degrees
=> Measurement at 0 degrees will consistently return the same value of M, measurement at 45 degrees returns 50/50 random result

* Send test photon polarized at 45 degrees.
=> Wavefunction collapses, one of the detectors is triggered, M is set to 0 or 1 with 50/50 probability as per Born rule
* Send a few vertically polarized readout photons, measure the result at 0 and 45 degrees
=> Measurement at 0 degrees will consistently return the same value of M, measurement at 45 degrees returns 50/50 random result

BTW when I say photons, I mean any suitable qubits will do as long as the governing logic is the same.

Now imagine we pack all this circuitry on a speck of silicon dust and somehow magically isolate it completely from the external environment (suspend it in vacuum inside superconducting chamber at 0K etc).

* Send test photon polarized at 45 degrees (or 135 or circular, the phase doesn't matter).
=> Device wavefunction splits into superposition of two branches, one branch has M=0, another M=1
=> Decoherence ensures the branches quickly diverge and do not interfere with each other
=> Within each branch "from the point of view of memory element M" wavefunction collapses and only one detector is triggered

* Send a few vertically polarized readout photons, measure the result at 45 degrees
=> photon gets entangled with the device and ends up in superposition of 0 and 90 degrees with equal amplitudes, that is, polarized at 45 degrees
=> measurement returns 1 with probability >50%, indicating the presence of superposition
=> measurement at 45 degrees does not disclose the value of M and so does not break the superposition

* Send vertically polarized readout photon, this time measure it at 0 degrees
=> the measurement randomly selects one branch and breaks the superposition
=> Subsequent measurements at 0 degrees will consistently return the same result
=> Subsequent measurements at 45 degrees will return 50/50 random result

Obviously there are enormous technical challenges implementing this, but conceptually, did I miss anything?

13. Jul 21, 2013

### Maui

David Deutsch is getting carried away. In Medieval times, it was thought by some that the spreading of deadly plague among people was caused by the evil powers of witches and this resulted in the infamous witch-hunts. It'd be interesting to see how long it will take till someone decides to blame unfortunate universe splitting for his car crash.

14. Jul 21, 2013

### Gerinski

Well I guess part of the question is, regarding superposition (which I guess nobody questions is an experimentally proven reality), do you consider each of the states as 'superpositioning states in our universe' or as 'states in superpositioning universes'? Maybe there's not much difference between both views, David Deutsch just prefers the second (I'm just quoting him).

And, what about his argument regarding quantum computation?

15. Jul 21, 2013

### stevendaryl

Staff Emeritus
That's unwarranted snideness, I think.

16. Jul 21, 2013

### Maui

We need to first define what we mean by our universe in quantum mechanical terms(impossible task these days) and then move on to err... other universes.

The only thing we can state from the POV of QM is what and how the universe is not. The rest is fantasy and I guess it's clear to everyone engaged with quantum theory.

At this stage, his comments about using other universes quantum states for quantum computing seem(to me) laughable and rest on a not so elegant, partial 'solution' to the measurement problem in qm.

I have no idea how he imagines putting and maintaining other universes in superposition given the obstacles engineering faces today with just a handful of atoms(and what would the inhabitants of those universes think of the idea), but then again I could be too conservative :tongue:

17. Jul 21, 2013

### Maui

Did you read the part about using other universes for better computational power?

18. Jul 21, 2013

### audioloop

a sine qua non condition, enough space.....

Topology of Branching Universes
http://arxiv.org/pdf/gr-qc/0505150.pdf
http://link.springer.com/article/10.1007/s10702-005-1319-9

in Maximal Branchs Universes there are logical problems
"tacitly supposes that there is a unique global time coordinate for the universe. Treating a measurement-like interaction as a point event in space-time, there will be many spacelike hypersurfaces which pass through that point; the selection of only one of these as the branching hypersurface requires one to accept that there is a preferential time coordinate for the universe."

Invariance Principle is lost.

in Minimal Branching Universes not.

---
and respect to the Empirical Viability

Against the Empirical Viability of the Deutsch-Wallace-Everett Approach to Quantum Mechanics
http://philsci-archive.pitt.edu/9541/1/Empirical_Viability_Archive.pdf

Second Testing

Excited-state decay in strictly Everett-like interpretations of quantum mechanics
http://arxiv.org/pdf/1209.3445v1.pdf
"if future work were to accurately determine a non-zero value of ε , this would constitute experimental conrmation of a prediction of strictly Everett-like formulations."

Experimentally testable geometric phase of sequences of Everett’s relative quantum states
http://arxiv.org/pdf/0903.1564v2.pdf
http://iopscience.iop.org/0295-5075/86/3/30005
"We have demonstrated that the geometric phase of sequences of relative quantum states can be tested in multiparticle interferometry, by means of local projective measurements and classical communication assisted postselection. We have argued that a mixed state generalization based on the Uhlmann holonomy leads to a multidimensional holonomy that depends on both classical and quantum correlations in a noisy bipartite system."

Last edited: Jul 21, 2013
19. Jul 21, 2013

### Dali

Actually, I do remember reading a preprint a long time ago that proposed a test that would do just that. The idea is pretty weird, but maybe interesting from a philosophical point of view...

The setup was that you connect a gun to a detector in such a way that it has a 50% chance of firing, and then you aim the gun at yourself! NB - do NOT try this at home! :-) The idea was then that if/when the universe splits into two copies - one where you die and another one where you live - you will only be aware of the one where you live. Thus, from your own perspective, you will be lucky every time you try this. You wouldn't be able to prove anything to anyone else though, as you would leave behind a lot of worlds where you did die.

I'm guessing (hoping?) the paper never got published. But it is at least a fun topic for discussions! Can't remember the author(s) now, or find any links. (It was probably 15 or so year ago I read this.)

Edit - apparently this setup is known as the "Quantum suicide", and wikipedia has an article about it: https://en.wikipedia.org/wiki/Quantum_suicide_and_immortality

20. Jul 23, 2013

### Maui

http://physicsworld.com/cws/article/news/2010/mar/18/quantum-effect-spotted-in-a-visible-object

According to this experiement, some universes will experience macroscopic superpositions without a cause and will behave nonlogically and inconsistently. If you suddeny observe something disappear, maybe somebody is bringing it to its ground state from another universe ;)

This is all of course assuming that the wavefunction is a real entity. Otherwise the future object to be put into superposition is not real and has no universe yet.

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook