Thread: The threshold for superposition collapse in double-slit experiment

What exactly is the threshold for collapse of superposition in a double-slit experiment? It is usually written that "measurement" or "observation" collapses the superposition. That is incredibly vague. The confusion rises when "intelligent observer" is mentioned as the cause for the collapse. So, has this been tested and what was the result?

Can a photon or wavefront traveling through a port be detected with absolutely zero change in the photon (except from the potential collapsing from the superposition)? Whether it can or not, what is the least interfering measurement possible, that will destroy the superposition?

Originally Posted by jimhoyle

What exactly is the threshold for collapse of superposition in a double-slit experiment? It is usually written that "measurement" or "observation" collapses the superposition. That is incredibly vague. The confusion rises when "intelligent observer" is mentioned as the cause for the collapse. So, has this been tested and what was the result?

In the many-worlds interpretation, there is no "collapse" of a superposition. Each orthogonal component of the superposition becomes quantum entangled with the distinct macroscopic states of the measuring device. In turn, each distinct macroscopic state of the measuring device becomes quantum entangled with distinct macroscopic states of the observer. Because a distinct macroscopic state of the observer is quantum entangled with a particular orthogonal component of the superposition, the observer only sees one outcome from the measurement of a quantum state, and this outcome is purely random. Therefore, the observer plays a crucial role in the apparent collapse of the superposition. However, it should be noted that the observer's role is passive, merely providing a first-person perspective of the measurement.

It is ultimately the orthogonality of the distinct components of a superposition, and the linearity of quantum mechanics, that is responsible for the apparent collapse of a quantum superposition because orthogonal states do not exhibit interference, and therefore behave as if the other components of the superposition do not exist.

Originally Posted by jimhoyle

Can a photon or wavefront traveling through a port be detected with absolutely zero change in the photon (except from the potential collapsing from the superposition)? Whether it can or not, what is the least interfering measurement possible, that will destroy the superposition?

If two photons are quantum entangled with each other, then a measurement of one photon is a measurement of the other photon even though there has been no physical interaction with this other photon.

Thank you for your good reply. Can it be said, for example in the case of double-slit experiment, in the many-worlds interpretation a chain of apparent superposition collapses (a kind of route of collapsing quantum states progressing through time) ends at the observer? This is at the heart of my question. So in an objective reality (not from the perspective of e.g. a human being but from the perspective of the universe or "many-worlds", for example) there is just an almost infinite chain of quantum states branching to almost infinite outcomes theoretically happening throughout the lifetime of the universe (which is infinite or finite). None of the outcomes are out of set laws of physics. For example in our universe all the outcomes would have galaxies, stars and planets. In even more objective reality the laws of physics may differ from many-worlds to other many-worlds for example, but that is another question.

But I am still not certain what happens in the actual double-slit experiment. If a human being perceives one outcome, then for that person there is an apparent collapse of the superposition. So if I put a detector on the ports and look at my detector monitor, I will see no interference pattern in the photon target screen. From my view, in my particular reality and moment in time the superposition has apparently collapsed at the port. But now comes my question again, in another words: if I put a detector on the ports but nobody ever turns the detector monitor on, will I still see no interference pattern (just because there was the detector)?

I assume I will not see the intereference pattern then (superposition still collapsed at the port). I assume if I turn my detector completely off, I will start to see the interference pattern. Sounds a bit like the Schrödinger's cat, but more straightforward. I would like to know what simply happens in the actual double-slit experiment.

Humans are nothing special in quantum mechanics. We're part of the same series of quantum states. We just can't perceive that for a number of reasons.

Originally Posted by jimhoyle

Thank you for your good reply. Can it be said, for example in the case of double-slit experiment, in the many-worlds interpretation a chain of apparent superposition collapses (a kind of route of collapsing quantum states progressing through time) ends at the observer?

No. From the perspective of an observer, a measurement acts to exclude entire histories and futures of entire realities. For example, if an experimenter prepares a superposition of two states, then later measures this to be in one of those states, then the prepared state was never in the other state. Only the version of the experimenter who observes the other state has that other state within his reality. But if the prepared superposition is never measured, then the superposition itself exists within the reality of the experimenter.

Originally Posted by jimhoyle

This is at the heart of my question. So in an objective reality (not from the perspective of e.g. a human being but from the perspective of the universe or "many-worlds", for example) there is just an almost infinite chain of quantum states branching to almost infinite outcomes theoretically happening throughout the lifetime of the universe (which is infinite or finite).

The universal wavefuntion evolves unitarily (deterministically). There is no objective branching occurring. There are an infinite number of versions of the one observer with the exact same history, and if a measurement is performed, this set is partitioned into sets of versions of the observer who observe each outcome. Each partition still contains an infinite number of versions of the observer, but the relative measure of each partition is proportional to the probability of the corresponding outcome. And because there is absolutely no way for the subjective observer to know which partition in which they belong, the particular observed outcome is purely random.

Originally Posted by jimhoyle

But I am still not certain what happens in the actual double-slit experiment. If a human being perceives one outcome, then for that person there is an apparent collapse of the superposition. So if I put a detector on the ports and look at my detector monitor, I will see no interference pattern in the photon target screen. From my view, in my particular reality and moment in time the superposition has apparently collapsed at the port. But now comes my question again, in another words: if I put a detector on the ports but nobody ever turns the detector monitor on, will I still see no interference pattern (just because there was the detector)?

I assume I will not see the intereference pattern then (superposition still collapsed at the port). I assume if I turn my detector completely off, I will start to see the interference pattern. Sounds a bit like the Schrödinger's cat, but more straightforward. I would like to know what simply happens in the actual double-slit experiment.

Ultimately, whether or not there is an interference pattern in the double-slit experiment depends on whether the components of the wavefunction from each slit are orthogonal or not.

Originally Posted by KJW

Originally Posted by jimhoyle

Thank you for your good reply. Can it be said, for example in the case of double-slit experiment, in the many-worlds interpretation a chain of apparent superposition collapses (a kind of route of collapsing quantum states progressing through time) ends at the observer?

No. From the perspective of an observer, a measurement acts to exclude entire histories and futures of entire realities. For example, if an experimenter prepares a superposition of two states, then later measures this to be in one of those states, then the prepared state was never in the other state. Only the version of the experimenter who observes the other state has that other state within his reality. But if the prepared superposition is never measured, then the superposition itself exists within the reality of the experimenter.

Originally Posted by jimhoyle

This is at the heart of my question. So in an objective reality (not from the perspective of e.g. a human being but from the perspective of the universe or "many-worlds", for example) there is just an almost infinite chain of quantum states branching to almost infinite outcomes theoretically happening throughout the lifetime of the universe (which is infinite or finite).

The universal wavefuntion evolves unitarily (deterministically). There is no objective branching occurring. There are an infinite number of versions of the one observer with the exact same history, and if a measurement is performed, this set is partitioned into sets of versions of the observer who observe each outcome. Each partition still contains an infinite number of versions of the observer, but the relative measure of each partition is proportional to the probability of the corresponding outcome. And because there is absolutely no way for the subjective observer to know which partition in which they belong, the particular observed outcome is purely random.

Originally Posted by jimhoyle

But I am still not certain what happens in the actual double-slit experiment. If a human being perceives one outcome, then for that person there is an apparent collapse of the superposition. So if I put a detector on the ports and look at my detector monitor, I will see no interference pattern in the photon target screen. From my view, in my particular reality and moment in time the superposition has apparently collapsed at the port. But now comes my question again, in another words: if I put a detector on the ports but nobody ever turns the detector monitor on, will I still see no interference pattern (just because there was the detector)?

I assume I will not see the intereference pattern then (superposition still collapsed at the port). I assume if I turn my detector completely off, I will start to see the interference pattern. Sounds a bit like the Schrödinger's cat, but more straightforward. I would like to know what simply happens in the actual double-slit experiment.

Ultimately, whether or not there is an interference pattern in the double-slit experiment depends on whether the components of the wavefunction from each slit are orthogonal or not.

That is a much more insightful and accessible articulation of these profound ideas than are to be found in any textbook or popularisation I've read. Beautifully done, KJW.

Originally Posted by KJW

Ultimately, whether or not there is an interference pattern in the double-slit experiment depends on whether the components of the wavefunction from each slit are orthogonal or not.

Very interesting this is. While I think about what you wrote, could you explain what you mean by the orthogonality here? Can you give me examples of cases when there is interference pattern and when not, regarding different kinds of types of measuring at the slits? What happens when simply tested in practice (the example I had)?

Originally Posted by jimhoyle

Originally Posted by KJW

Ultimately, whether or not there is an interference pattern in the double-slit experiment depends on whether the components of the wavefunction from each slit are orthogonal or not.

Very interesting this is. While I think about what you wrote, could you explain what you mean by the orthogonality here? Can you give me examples of cases when there is interference pattern and when not, regarding different kinds of types of measuring at the slits? What happens when simply tested in practice (the example I had)?

I think the following example is very instructive:

Consider the case that there is a device behind one of the slits that rotates the plane of polarisation by 90°. Because the component states from each slit are now orthogonal, there is no interference. The experimenter can determine which slit each photon passed through by measuring the polarisation of each photon (assuming the photons passing through the slits are plane-polarised to begin with). But even if the observer does not measure the polarisation of the photons, there is still no interference because the component states from each slit are orthogonal.

One way to think about measurement is not to consider this photon state as being orthogonal to that photon state, but rather to consider this entire reality containing this photon state as being orthogonal to that entire reality containing that photon state, with the question of interference or non-interference being with regards to the entire realities. Then when one uses a macroscopic device to measure which slit the photon passed through, it's the distinct states of the macroscopic device that are orthogonal because of the large number of atoms that form the distinct states, and these distinct macroscopic states do not interfere. But because the distinct photon states belong to the same realities as the corresponding distinct macroscopic states, the non-interference of distinct macroscopic states manifests as non-interference of distinct photon states.

Originally Posted by KJW

I think the following example is very instructive: ...

Agreed, a good example! Makes me think more deeply about actual experimental setups:

For a two-slit experiment with photons, when considering versions involving single photons, is there some relationship between "wavelength" (of a single photon?) and the spacing of the two slits, before interference is possible?

Then the electron case. Given that business in 2012 (Swiss Light Source) of teasing apart certain properties of a confined electron (holon, spinon, orbiton), the idea of electron as extended object (something other than a mere statistical blip) becomes interesting. When modelling possible models for an electron subject to double slit interactions, to be consistent with the data, how close and how wide do those two slits need to be?

Also, do real physicists still entertain alternatives to a "many-worlds" world view?

Originally Posted by jimhoyle

What exactly is the threshold for collapse of superposition in a double-slit experiment? It is usually written that "measurement" or "observation" collapses the superposition. That is incredibly vague. The confusion rises when "intelligent observer" is mentioned as the cause for the collapse. So, has this been tested and what was the result?

Can a photon or wavefront traveling through a port be detected with absolutely zero change in the photon (except from the potential collapsing from the superposition)? Whether it can or not, what is the least interfering measurement possible, that will destroy the superposition?

There is no such thing as superposition. this is because the photons will be affected by other photons, as, photons are everywhere, as, they are massless. even in a dark room, there are photons that and other things that will affect the stream.

But, if two photons are fired at the same place, well, it could be a million photons, how many are there inside the stream it could be asked, they could occupy the same space and have the same effects done to them.

Now, if they are fired from different places, they will encounter different things. if the things they encounter are similar, they still won't be the same. being massless, they will go right through these things, you could say, but, they are still somewhat affected by things that affect photons, all of which we are not aware of yet.

Originally Posted by Brett

There is no such thing as superposition. ...{wrong and muddled explanation excised}

The equations of QM are linear. Superposition holds generally.