Experimental Metaphysics

[pood]

Interesting (about 5,000 words) article from quanta magazine about “experimental metaphysics.”

The term was coined in 1980, though the author maintains it originated with Einstein’s friend Michael Besso. It refers to the first experimental test of Bell’s Inequality, an experiment that vindicated quantum mechanics against Einstein’s objections. The outcome of the experiment showed that one cannot maintain BOTH “locality” (no “spooky action at a distance,” per Einstein) AND realism (the notion of a mind-independent reality). One of the other has to go.

It’s not that simple, though. The article failed to mention that one CAN maintain both locality and realism, and in addition determinism as against quantum indeterminism, if one adopts the many worlds interpretation. It does mention MW later, briefly, in a different context.

Also, the experimental results are consistent with both locality and realism under superdeterminism, which denies statistical independence, the idea that there is independence between past hidden variables and current experimental settings. The article later explains superdeterminism as a “cosmic conspiracy tricking you into setting your detectors so that the outcomes seem to violate Bell’s inequality even though they don’t.” This outcome is obtained by pre-determining the fallacious choice of detector settings at the start of the universe.

More broadly, the article contends that metaphysics and science constitute an inseparable whole, and one can’t have one without the other, This would irk a lot of scientists who, like Steven Hawking, think philosophy is useless.

It gives a nice example of how Einstein’s relativity appeared to rule out Euclidean space, but of how Poincaré responded with a thought experiment involving keeping Euclidean space and adding a combination of temperature gradients and refraction that would make the universe appear to be non-Euclidean while actual maintaining a Euclidean structure.

This is theory underdetermination, the idea that there could be potentially an infinite number of theories to explain the same set of data provided one adjusts auxiliary assumptions, and seriously undermines putting science on a firm metaphysical footing —severally compromising, for example, Popperian falsification.

There is a lot more in the article, including an extended discussion of Wigner’s Friend and later variants that put the idea of objective, mind-independent reality under serious question.

I put the article out there for possible discussion in case others, like me, are sick of talking about politics and the Orange Monster From MAGGOTland.

 

[Swammerdami]

I've not finished the article yet, but want to muse on Doesn't the GHZ experiment offer a crisper look at so-called "spookiness" than Bell's Theorem?

Bell is over-rated! Oh, there's no doubt that "Bell's Theorem", when taken as an umbrella term for types of quantum spookiness, is very important, but the actual Theorem is over-rated.

Bell's result was a simple corollary of basic quantum theory, as was already well known (e,g, David Bohm (1952)). The experiments, using two entangled particles, which demonstrated Bell's result had predictable results.

Why not substitute the GHZ experiment (which uses THREE entangled particles) to contemplate the "spookiness"? It doesn't require any probabilistic analysis; instead an alleged hidden variable changes state absolutely. I will describe the GHZ experiment as an elaborate version of Three-Card Monte. (With sufficient interest I'll hunt down a good URL for GHZ.)

. . . . . . . .

The dealer repeatedly presents you with three objects. Rearrange them all you want: it doesn't matter. Each object has two windows,, either of which you can open to reveal a color; but once you open one window the other is inaccessible.

Opening the green window will present you with one of two colors: green or orangish red. Similarly opening the blue window will present you with blue or orangish yellow. One window is opened on each object, so three colors altogether will appear. Observe the deals as long as you like; the results always fit a pattern: When zero or two green windows are open, the eight possible color combos each appear 1/8 of the time. But when exactly one green window is opened, the number of orangish colors will always be an odd number. When all three green windows are opened, the number of orangish colors will always be an even number.

Now let's look at one specific trial. We'll assume you start by seeing two greens, but the result will be similar for any combination.

[I intended to use the TABLE tag for this presentation but discover that this board's implementation of TABLE is utterly sadistic.]

Here is the state after the two openings. "XXXXX" means that we will never be able to open that window.

Behind Green Window: WWWWWWWWWWWWWW...WWWWWWWWWWWWW... .. Green! ..
Behind Blue. Window: ... XXXXX .......... XXXXX ......... ????? ...

We've opened the Green window on objects #1 and #2 and seen green and green. Under the rules ("When all three green windows are opened, the number of orangish colors will always be an even number.") we know that the third green window will present green so there's no point in opening that window; we open the blue window on object #3. Suppose we see blue inside:

Behind Green Window: GGGGGGGGGGGGGG...GGGGGGGGGGGGGG...gggggggggggggg
Behind Blue. Window: ... XXXXX .......... XXXXX ......BBBBBBBBBBBBBB

I've used a paler green to show the color behind the green window on object #3: We've not actually SEEN that color; we've DEDUCED it.

Now suppose that we'd opened the green window on object #1, and blue window on object #3 as before, but the blue window on object #2. Under the rules ("when exactly one green window is opened, the number of orangish colors will always be an odd number.") we can deduce that that orange will be revealed:

Behind Green Window: GGGGGGGGGGGGGG...GGGGGGGGGGGGGG...gggggggggggggg
Behind Blue. Window: oooooooooooooo...OOOOOOOOOOOOOO...BBBBBBBBBBBBBB

But that's impossible! We'd see Orange/Orange/Green if we'd opened Blue/Blue/Green windows respectively, VIOLATING the rule ("when exactly one green window is opened, the number of orangish colors will always be an odd number.")

I think I could have rendered a better argument at a blackboard, but editing as I did was laborious. You are welcome to work out other cases.
I prefer GHZ compared with the usual Bell's theorem experiment because NO probabilities are involved (nor cosines!) We PROVE that a hidden color MUST be green, yet also prove that it MUST be orange! Object #3 had to "know" what happened (or what WILL happen) with the other objects before it can "decide" what color to present.

Solution to this and similar spookiness is easy: Just reverse the cause/effect arrow connecting the objects!

 

[pood]

GHZ experiment

It is noted that the experiment yet again agrees with quantum mechanics.

 

[pood]

Bell Inequality, GHZ and the Many Worlds

The upshot is that the Many Worlds Interpretation removes all the riddles of QM. There is no wave-function collapse, no spooky action at a distance, no anti-realism and no indeterminism

 

[steve_bank]

Experimental metaphysics has no meaning.

Experimental science is proving an idea with an experiment.

So far there is no way to determine experimentaly if the stats cal nature of reality represents uncertainty er not.

When uranium emits a particle we can only assigned a probability. Is that an artifact or measurement ,or is the universe deterministic? Subjective philosophical interpretation of experiment. Speculative m,metaphysics and philosophy.

BTW, the existence of god answers all questions and resolves all problems, does it not?

 

[pood]

Experimental metaphysics has no meaning.

Experimental science is proving an idea with an experiment.

So far there is no way to determine experimentaly if the stats cal nature of reality represents uncertainty er not.

When uranium emits a particle we can only assigned a probability. Is that an artifact or measurement ,or is the universe deterministic? Subjective philosophical interpretation of experiment. Speculative m,metaphysics and philosophy.

BTW, the existence of god answers all questions and resolves all problems, does it not?

The point the article is making is that QM has a number of different metaphysical interpretations that could never be tested. Then, in 1980, it was possible to do a direct experimental test of the Bell’s inequality, and the result turned out to confirm the “spooky action at a distance” that Einstein rejected with his earlier thought experiment. But even so, that result can be challenged by two other metaphysical interpretations — superdeterminism or Many Worlds. The larger point the article is making is that science and metaphysics are inextricably intertwined, a point reinforced by the Poincaré thought experiment that showed how Einstein’s curved spacetime could be interpreted in a different way that restored Euclidean space.

 

[Bomb#20]

Bell Inequality, GHZ and the Many Worlds

The upshot is that the Many Worlds Interpretation removes all the riddles of QM.

Well, except the big one: how do events in infinite-dimensional configuration space cause effects in ordinary three-D space? ...

There is no wave-function collapse,

... Oh, and the extra idiosyncratic riddle MWI throws into QM's mix: why are high-amplitude outcomes observed more frequently than low-amplitude outcomes? (Wave-function collapse makes that happen in more conventional interpretations.)

no spooky action at a distance, no anti-realism and no indeterminism

"Alice first splits the world into four different worlds according to her measurements. Each of the worlds is then split again into four worlds by Bob. Charley, however, does not make any additional splitting. In every one of the 16 worlds created by Alice and Bob, the outcomes of his two spin measurements are already fixed."​

So if Bob hasn't already made his measurement, Charley's measurement splits the world, but if he has, it doesn't, even if Bob is ten feet away and even if he makes his measurement one nanosecond before Charley's. Whether Charley's act of measurement causes the world to split depends on an event outside its light-cone. Sure as heck looks to me like spooky action at a distance.

 

[pood]

Bell Inequality, GHZ and the Many Worlds

So if Bob hasn't already made his measurement, Charley's measurement splits the world, but if he has, it doesn't, even if Bob is ten feet away and even if he makes his measurement one nanosecond before Charley's. Whether Charley's act of measurement causes the world to split depends on an event outside its light-cone. Sure as heck looks to me like spooky action at a distance.

Just to answer this one point; I’ll address the rest later.

Yes, “spooky action at a distance” happens outside the light cone under wave-function collapse interpretations, but this does not violate special relativity’s ban on faster-than-light signaling, because no information can be exchanged outside the light cone. The point of the many worlds, as the author of an article I linked above explains, is that there simply is no “spooky action at a distance” under MWI. There is, rather, a zero quantum probability of living in any world in which particles outside the light cone do not appear to be quantum entangled.

 

[steve_bank]

I am, what Popper defined as an experimentalist meaning all we really know for sure is the results of a controlled ex[merriment. It is the only thing we can call objective knowledge.

As debate and discussion expands around experiment it becomes less objective and more subjective.

The multi universe QM interpretation is speculation because it can not be experimentally demonstrated.

When String Theory came out some considered it philosophy not science.

In the past on the forum some claimed science is really philosophy, I disagree.

Experimental psychology is based in controlled experiment

Using demonstrated QM as the basis for a metaphysical creation is not an excremental demonstration.

The word metaphysics is a muddy term. To me it means abstract thought constructs not tied to physical reality. Metaphysics has been superseded by modern empirical science and psychology.

Metaphysics (Stanford Encyclopedia of Philosophy)

 

[pood]

Using demonstrated QM as the basis for a metaphysical creation is not an excremental demonstration.

I should hope not!

But, kidding aside, as mentioned, I think the linked article makes a pretty powerful argument that science and metaphysics are inextricably bound up together. As the philosopher Norman Swartz noted, scientists who disparage philosophy seem unaware that science itself is shot through with metaphysical assumptions which themselves cannot be proved or falsified.

 

[Bomb#20]

Bell Inequality, GHZ and the Many Worlds

So if Bob hasn't already made his measurement, Charley's measurement splits the world, but if he has, it doesn't, even if Bob is ten feet away and even if he makes his measurement one nanosecond before Charley's. Whether Charley's act of measurement causes the world to split depends on an event outside its light-cone. Sure as heck looks to me like spooky action at a distance.

Just to answer this one point; I’ll address the rest later.

Yes, “spooky action at a distance” happens outside the light cone under wave-function collapse interpretations, but this does not violate special relativity’s ban on faster-than-light signaling, because no information can be exchanged outside the light cone. The point of the many worlds, as the author of an article I linked above explains, is that there simply is no “spooky action at a distance” under MWI. There is, rather, a zero quantum probability of living in any world in which particles outside the light cone do not appear to be quantum entangled.

Sorry if I'm being obtuse, but I read the article, and if there's a place in the article that explains why causing a particle to be spin-up that would otherwise have been spin-down counts as an "action", but preventing a universe split that would otherwise have occurred does not count as an "action", I didn't see it.

 

[pood]

Bell Inequality, GHZ and the Many Worlds

So if Bob hasn't already made his measurement, Charley's measurement splits the world, but if he has, it doesn't, even if Bob is ten feet away and even if he makes his measurement one nanosecond before Charley's. Whether Charley's act of measurement causes the world to split depends on an event outside its light-cone. Sure as heck looks to me like spooky action at a distance.

Just to answer this one point; I’ll address the rest later.

Yes, “spooky action at a distance” happens outside the light cone under wave-function collapse interpretations, but this does not violate special relativity’s ban on faster-than-light signaling, because no information can be exchanged outside the light cone. The point of the many worlds, as the author of an article I linked above explains, is that there simply is no “spooky action at a distance” under MWI. There is, rather, a zero quantum probability of living in any world in which particles outside the light cone do not appear to be quantum entangled.

Sorry if I'm being obtuse, but I read the article, and if there's a place in the article that explains why causing a particle to be spin-up that would otherwise have been spin-down counts as an "action", but preventing a universe split that would otherwise have occurred does not count as an "action", I didn't see it.

I’m not sure what your objection is here. Perhaps you could clarify. Under MWI, there is no spooky action at a distance. That only exists under the assumption of wave-function collapse.

 

[Bomb#20]

... The outcome of the experiment showed that one cannot maintain BOTH “locality” (no “spooky action at a distance,” per Einstein) AND realism (the notion of a mind-independent reality). One of the other has to go.

It’s not that simple, though. The article failed to mention that one CAN maintain both locality and realism, and in addition determinism as against quantum indeterminism, if one adopts the many worlds interpretation. ...

Bell Inequality, GHZ and the Many Worlds

The upshot is that the Many Worlds Interpretation removes all the riddles of QM. There is no wave-function collapse, no spooky action at a distance, no anti-realism and no indeterminism

Although there is no action at a distance in the MWI, it still has nonlocality.

You and Dr. Vaidman are using conflicting definitions of either "action at a distance" or "locality".

 

[pood]

... The outcome of the experiment showed that one cannot maintain BOTH “locality” (no “spooky action at a distance,” per Einstein) AND realism (the notion of a mind-independent reality). One of the other has to go.

It’s not that simple, though. The article failed to mention that one CAN maintain both locality and realism, and in addition determinism as against quantum indeterminism, if one adopts the many worlds interpretation. ...

Bell Inequality, GHZ and the Many Worlds

The upshot is that the Many Worlds Interpretation removes all the riddles of QM. There is no wave-function collapse, no spooky action at a distance, no anti-realism and no indeterminism

Although there is no action at a distance in the MWI, it still has nonlocality.

You and Dr. Vaidman are using conflicting definitions of either "action at a distance" or "locality".

How so?

 

[Bomb#20]

... The outcome of the experiment showed that one cannot maintain BOTH “locality” (no “spooky action at a distance,” per Einstein) AND realism (the notion of a mind-independent reality). One of the other has to go.

It’s not that simple, though. The article failed to mention that one CAN maintain both locality and realism, and in addition determinism as against quantum indeterminism, if one adopts the many worlds interpretation. ...

Bell Inequality, GHZ and the Many Worlds

The upshot is that the Many Worlds Interpretation removes all the riddles of QM. There is no wave-function collapse, no spooky action at a distance, no anti-realism and no indeterminism

Although there is no action at a distance in the MWI, it still has nonlocality.

You and Dr. Vaidman are using conflicting definitions of either "action at a distance" or "locality".

How so?

You are equating the two concepts. He is not. According to your usage, "locality" means "no spooky action at a distance". If Dr. Vaidman were using "locality" in that sense then "nonlocality" would imply "spooky action at a distance", and therefore "action at a distance. But he explicitly denies that implication.

 

[pood]

... The outcome of the experiment showed that one cannot maintain BOTH “locality” (no “spooky action at a distance,” per Einstein) AND realism (the notion of a mind-independent reality). One of the other has to go.

It’s not that simple, though. The article failed to mention that one CAN maintain both locality and realism, and in addition determinism as against quantum indeterminism, if one adopts the many worlds interpretation. ...

Bell Inequality, GHZ and the Many Worlds

The upshot is that the Many Worlds Interpretation removes all the riddles of QM. There is no wave-function collapse, no spooky action at a distance, no anti-realism and no indeterminism

Although there is no action at a distance in the MWI, it still has nonlocality.

You and Dr. Vaidman are using conflicting definitions of either "action at a distance" or "locality".

How so?

You are equating the two concepts. He is not. According to your usage, "locality" means "no spooky action at a distance". If Dr. Vaidman were using "locality" in that sense then "nonlocality" would imply "spooky action at a distance", and therefore "action at a distance. But he explicitly denies that implication.

I may have worded it poorly but what he is saying is that under MWI, there is entanglement but there is no spooky action at a distance.

 

[Bomb#20]

So if Bob hasn't already made his measurement, Charley's measurement splits the world, but if he has, it doesn't, even if Bob is ten feet away and even if he makes his measurement one nanosecond before Charley's. Whether Charley's act of measurement causes the world to split depends on an event outside its light-cone. Sure as heck looks to me like spooky action at a distance.

Just to answer this one point; I’ll address the rest later.

Yes, “spooky action at a distance” happens outside the light cone under wave-function collapse interpretations, but this does not violate special relativity’s ban on faster-than-light signaling, because no information can be exchanged outside the light cone. The point of the many worlds, as the author of an article I linked above explains, is that there simply is no “spooky action at a distance” under MWI. There is, rather, a zero quantum probability of living in any world in which particles outside the light cone do not appear to be quantum entangled.

Sorry if I'm being obtuse, but I read the article, and if there's a place in the article that explains why causing a particle to be spin-up that would otherwise have been spin-down counts as an "action", but preventing a universe split that would otherwise have occurred does not count as an "action", I didn't see it.

I’m not sure what your objection is here. Perhaps you could clarify. Under MWI, there is no spooky action at a distance. That only exists under the assumption of wave-function collapse.

"What I tell you three times is true." You keep asserting there is no spooky action at a distance in MWI, but without ever explaining it. When I was in college the philosophy department had a motto: "What does it mean? How do you know?". So what does "There is no spooky action at a distance in MWI." mean? And how do you know there is no spooky action at a distance in MWI?

Dr. Vaidman does no better. His explanation reads as follows:

"To see explicitly how the MWI removes the action at a distance of Bell-type experiments consider a demonstration of the GHZ experiment...
[technical details snipped]
As shown above, the MWI removes action at a distance from quantum physics."​

[My bold]

Note that the word "the" has disappeared between his argument and his conclusion. There is one specific action at a distance in conventional interpretations that Dr. Vaidman has shown MWI removes. But that is not the same thing as showing MWI removes action at a distance in general from quantum physics. To all appearances, switching to Many Worlds merely replaces one action at a distance with a different action at a distance.

Vaidman continues,

"Like in classical relativistic physics, any local action on a system changes nothing whatsoever at remote locations at the moment of disturbance."​

Taken literally, that's a nonsense expression -- in classical relativistic physics, "at remote locations at the moment of disturbance" is a meaningless phrase because classical relativistic physics doesn't have absolute simultaneity. Presumably he meant it nonliterally, as a stand-in for "Like in classical relativistic physics, any local action on a system changes nothing whatsoever at remote locations with space-like separation from the disturbance event." But, as I pointed out upthread, that simply isn't true. In MWI, Bob's local act of disturbance changes whether Charlie's measurement splits the universe in two or not.

So my objection here is, how the heck does splitting the universe in two qualify as "nothing whatsoever"?

 

[pood]

In non-MWI QM, the particles are entangled and the observation of one particle “collapses” the wavefunction to instantaneously dictate the state of the other particle even if it is on the other side of the universe. However, this does not violate SR since the phenomenon cannot be exploited for superluminal signaling.

In MWI, there is no wavefunction collapse, so the observation of one particle cannot affect the state of the other. Each observer measures one particle and each measurement device birfurcates into two versions, one for each possible measurement outcome. The correlations of the separated particles are later established locally by comparison. These correlations remain the same in MWI as in non-MWI, but there is no causation as there is in non-MWI. An illustration of why correlation is not causation, as the old saw goes.