Why Would Observation Change Anything Physically?

[ray3400]

TL;DR Summary

Question about the cause and effect relationship between observation and physical particles changing.

A particle has a 33% chance of being in either position 1, position 2, or position 3. After I observe it, it is in position 1, and not in position 2 or 3.

Questions:
How do we know it was not already in that position prior to us observing it? Does observation cause position, or is position absolute and acknowledged by observation?

Why would me simply observing a particle determine its position?
To say that observing something can cause it to change positions or to solidify itself into a single position is the same logic as saying "by the act of me looking at a door, the door opened." We know that in traditional physics there needs to be an exchange between systems - whether it be matter or energy - in order for one system to affect another. I.e. the molecules of my hand applied a force to the molecules of a door handle, accelerating the door so that it is in an open position. Observation alone does not imply an exchange of matter or energy between systems, so I don't understand how there can be a cause and effect relationship between observation and position.

 

[metastable]

A particle has a 33% chance of being in either position 1, position 2, or position 3. After I observe it, it is in position 1, and not in position 2 or 3.

Questions:
How do we know it was not already in that position prior to us observing it?

If the particle is a photon and the 3 positions are 3 people’s eyeballs, we know it wasn’t there prior to observing it because the speed of light appears the same in all rest frames.

Does observation cause position, or is position absolute and acknowledged by observation?

How would you measure the photon’s position unless it was seen? How would the same photon be seen by more than one person? If observation has no effect, can all 3 people take independent measurements of the same photon?

 

[A. Neumaier]

Observation alone does not imply an exchange of matter or energy between systems

It does involve at least an energy exchange. Though it is minute for large objects, it is (relatively) substantial for tiny objects.

 

[metastable]

If you consider running particles through successive stern-gerlach apparatuses to be “taking measurements,” it seems to affect the particles in some way...

 

[Nugatory]

Observation alone does not imply an exchange of matter or energy between systems, so I don't understand how there can be a cause and effect relationship between observation and position.

An “observation” requires some interaction with whatever is being observed. For example, you might sense the position of an object in a dark room by feeling for it; but you won’t feel it unless the flesh of your outstretched fingertip is very slightly compressed, meaning that the object is exerting a small force on your fingertip and by Newton’s second law your fingertip is exerting a small force on the object. The force is small, but it’s there and it will affect the object. Think about other ways of finding the position (echolocation like bats do, bouncing radar or light waves off it, looking for its shadow, ...) and you’ll see that they all involve some exchange of energy and momentum with the object, thereby changing it in some small way.

It’s worth noting that in quantum mechanics any thermodynamically irreversible interaction with a system counts as an “observation”. No observer need be involved; for example a book sitting on a table is being continuously “observed” by its interactions with the air molecules surrounding it. Thus the word doesn’t mean quite what it does in ordinary English; it’s a historical accident that physicists started using it this way more than a century ago.

How do we know it was not already in that position prior to us observing it? Does observation cause position, or is position absolute and acknowledged by observation?

If it’s a macroscopic object, we do know that it was in that position before we observed it. If we use the methods of quantum mechanics on a macroscopic object, we don’t get answers like “33% here, 33% there, 33% somewhere else”, we get something like “100% right there” - so of course we don’t bother with the more complicated and less intuitive quantum mechanical solution, we just use classical mechanics and our intuitive sense that objects have definite positions at all times. Quantum mechanical effects generally only show up with much smaller particles, but then they cannot be ignored. The double slit experiment (done with particles instead of waves) is one of the simplest examples: if the incoming particles had definite positions at all times, each particle would have to pass through one slit or the other, but that produces a different pattern than is actually observed.

 

[hutchphd]

It’s worth noting that in quantum mechanics any thermodynamically irreversible interaction with a system counts as an “observation”.

So the Stern-Gerlach is only "observing" if one of the channels is occluded? There is still an interaction...but reversible I think.

 

[PeroK]

Summary: Question about the cause and effect relationship between observation and physical particles changing.
A particle has a 33% chance of being in either position 1, position 2, or position 3.

This can better be rephrased:

If you measure the position of a particle, there is a 33% of obtaining position 1; a 33% chance of obtaining position 2; and, a 33% chance of obtaining position 3.

Until you measure the position of the particle, it does not have a definite position.

 

[metastable]

Can I say any measurement of an electron in a double slit experiment will affect the electron?

 

[PeroK]

Can I say any measurement of an electron in a double slit experiment will affect the electron?

In the Copenhagen interpretation, a measurement changes the state of the electron - to an eigenstate of the measured observable.

 

[metastable]

Does that mean the measurement changed the electron or the measurement revealed the state of the electron prior to the measurement?

 

[PeroK]

Does that mean the measurement changed the electron or the measurement revealed the state of the electron prior to the measurement?

A single measurement cannot reveal the state. This is where it is better to look at spin than position.

A measurenent of the spin of an electron (about a given axis) can only return $\pm \frac{\hbar}{2}$. And the spin state of the electron after the measurement is the appropriate eigenstate.

It doesn't matter what the spin state of the electron was before the measurement: you must get one of these two values. Only statistically, by repeated measurements on an ensemble of identically prepared particles can you deduce the initial state.

Note that the state of the electron before the measurement is not a definite "spin". In fact, quantum spin states fundamentally contradict the classical notion of a spinning particle having a definite spin in a definite direction.

There is no way to sustain the notion of a definite spin before measurement in the light of the Stern-Gerlach experiment, for example,

 

[metastable]

you must get one of these two values.

only 2 values but any spin vector & any precession or rate of change to the precession?

 

[PeroK]

only 2 values but any spin vector & any precession or rate of change to the precession?

Spin precession is an evolution of the state. This leads to evolving expectation values of spin measurements; which are still, always individually $\pm \frac{\hbar}{2}$.

 

[metastable]

always individually ±ℏ2±ℏ2\pm \frac{\hbar}{2}.

Wouldn’t measuring this value from each electron require measuring each individual electron along a different individual “custom tailored” axis?

 

[PeroK]

Wouldn’t measuring this value from each electron require measuring each individual electron along a different individual “custom tailored” axis?

I don't understand that question.

 

[metastable]

When measuring the spin value along the spin axis of each electron, how do we know this axis in advance to make the measurement? For every electron won’t we have to use a custom “measuring rod” that “rotates?” But I thought quantum particles can’t actually rotate in the classical sense...

 

[PeroK]

When measuring the spin value along the spin axis of each electron, how do we know this axis in advance to make the measurement? For every electron won’t we have to use a custom “measuring rod” for each electron that “rotates?” But I thought quantum particles can’t actually rotate in the classical sense...

It doesn't matter what axis you choose. That's the point. If you have a classical object that always spins at the same rate, then if you choose (by luck or prior knowledge) the axis of rotation, you'll get all the spin; and if you choose an axis perpendicular to this you'll get no spin.

But, with an electron you never get all the spin and you never get no spin about a chosen axis. Crudely, you always get 1/3 of the total spin; whatever axis you choose.

This is what is fundamentally incompatible with the classical concept of a spinning object.

 

[metastable]

The spin can only can have 2 values when measured along the axis, but can we agree the magnetic moment for each electron could be pointing in any direction, and each electron can have an individual precession rate, and an individual rate of change of precession?

 

[Nugatory]

So the Stern-Gerlach is only "observing" if one of the channels is occluded? There is still an interaction...but reversible I think.

How would you characterize (in an I-level thread and without opening a can of interpretational worms) the interactions that we treat as collapsing the wave function?

I'm asking not arguing here - I know that there's an element of lying to children in tossing out "thermodynamic irreversability" without qualification.

 

[PeroK]

The spin can only can have 2 values when measured along the axis, but can we agree the magnetic moment for each electron could be pointing in any direction, and each electron can have an individual precession rate, and an individual rate of change of precession?

An electron, in a given state, will have an expected value of spin about any axis. The expected value can take any value, between $\pm \frac{\hbar}{2}$. That's because you have a distribution of probabilities for the two values. If the probabilities are equal, then the expected value of spin is 0. That comes from a statistically equal number of plus and minus measurements.

If plus is more likely than minus, then you could have an expected value of $+\frac{\hbar}{6}$ or whatever.

This is analagous to the expected value of a loaded die could be anything from 1-6, although each individual measurement is a definite 1, 2, 3, 4, 5 or 6.

 

[metastable]

The expected value can take any value, between ±ℏ2±ℏ2\pm \frac{\hbar}{2}. That's because you have a distribution of probabilities for the two values.

What if the electron in question is entangled and I measured the other particle already?

 

[PeroK]

What if the electron in question is entangled and I measured the other particle already?

You're just being silly now!

 

[Nugatory]

The spin can only can have 2 values when measured along the axis, but can we agree the magnetic moment for each electron could be pointing in any direction, and each electron can have an individual precession rate, and an individual rate of change of precession?

We cannot. In quantum mechanics the only statements that everyone can agree about will be of the form "A measurement of X was performed and produced the result Y" or "If a measurement of X is performed, the probability that the result will be Y is ...".

 

[andresB]

I would like to point out to the OP the detection-free measurements
https://arxiv.org/abs/hep-th/9305002https://arxiv.org/abs/0801.2777

 

[ray3400]

Thanks everyone, there was a misunderstanding on my part about how observation works on the quantum level. I now understand how observing a quantum particle can affect the results of an experiment.