# I Question About The Role of Observation in Quantum Mechanics

#### PeroK

Homework Helper
Gold Member
2018 Award
Here is another study that explores the same question:
https://www.sciencedaily.com/releases/1998/02/980227055013.htm

"REHOVOT, Israel, February 26, 1998--One of the most bizarre premises of quantum theory, which has long fascinated philosophers and physicists alike, states that by the very act of watching, the observer affects the observed reality. "

"the very act of watching, the observer affects the observed reality"

How does the "observer effect" cause the two different output patterns in the double slit experiment for electrons? There is a different output pattern when the electrons are being observed before the going through the slits and when they are not being observed before going through the slits.
This is the sort of thing that you need to start ignoring if you are going to learn any QM.

There is a common experiment where you measure the speed of a bullet by firing it into block of wood and measure the momentum of the block after the collision. In this case, the measurement of the bullet's speed/momentum fairly brutally changes the speed of the bullet. This is clearly of no significance to the question here.

The fundamental point about the double-slit experiment is that if you do not measure which slit the electron passes through, then the question of which slit it passes through no longer has meaning.

In the Feynman lecture (which you really should watch), he sums it up as follows (in the case where thjere is no detector):

Proposition: the electron must either pass through slit A or pass through slit B.

He then shows that, using the experimental results of the double slit, that proposition fails. That proposition seems so close to basic logic that it is hard to understand how it could fail. But, the proposition fails, nevertheless. You cannot say that the electron passed through either slit A or slit B; and, you cannot say that the electron passed through both slits; and, you cannot say that the electron passed through neither slit. You cannot say the electron behaved like a particle; you cannot say the electron behaved like a wave; and you certainly cannot say that an electron behaved like a wave until it was observed and thereafter it behaved like a particle.

This is where it becomes fundamentally important to understand the electron not as a classical particle that has a well-defined trajectory, but as a quantum object whose position (if and when you measure its position) is governed by a probabilistic wave-function.

The electron never is anywhere or doing anything that is changed by an observation. That's the difference with the bullet and the block. The bullet, statistically at least, really has a well-defined position, momentum and trajectory which is changed by its collision with the block. It makes no sense to ask where the electron was if you didn't measure where it was.

One final point is that it is also worth studying the single-slit behaviour of quantum objects. Another lecture I would recommend in this respect is the following. The single-0slit experiment is analysed and carried out from about 32 mins in (although the whole lecture is worth watching if you want to understand QM).

#### PeterDonis

Mentor
In the experiment when the electrons are being observed using a detector they produce a different output pattern than when they are not being observed with a detector. What about the detector observing them causes the output pattern to change?
The fact that the detectors at the slits are physical systems that interact with the electrons and change their behavior. "Observation" is an interaction, and interaction changes things.

Exactly how they interact with the electrons and change their behavior is what you need the detailed math of QM for.

#### DrChinese

Gold Member
In the double-slit experiment when a detector was placed before the two slits, a 2 strip pattern was produced after the two slits. When there was no detector placed before the two slits, a different pattern was produced after the two slits. Why does the presence of a detector before the two slits cause a different pattern to be produced after the two slits?
Generally, no one can explain the "why" of physics when you are discussing fundamentals. And the fundamental rule for the double slit setup is: If it is *possible* to know which slit a particle goes through, then there will be NO interference. QM properly explains this without explaining "why" physics is like this in the first place. Of course the explanation involves a good deal of QM, here is an example using optics (photons instead of electrons or ions):

In several posts, you speculated that the detector itself causes the pattern to change. But that is not really the case, as you can see according per the rule above. And there are in fact experiments that demonstrate this. That is because you do NOT need a detector at all to control the presence or absence of interference. In an optical version of the experiment, you can place *parallel* polarizers at 45 degrees in front of the slits (like: / /) . There WILL be interference. So there must not be a detector present.

BUT... if you re-orient the polarizers so they are *crossed* (perpendicular, orthogonal, still at 45 degrees, like: \ /): there will NOT be interference. Same apparatus otherwise, and note that we have no idea which slit the photons go through. So what gives?

Going back to the rule above: When the polarizers are crossed, it is POSSIBLE to determine which slit the photon passes through even though we didn't attempt to learn that information. Therefore, no interference, and there is no detector present at all. So it can't be the detector.

#### atyy

Summary: Does observing a particle cause it to exhibit a certain quality? What is the cause and effect relationship involved with observation?
Observation does have a special role in quantum mechanics (using the orthodox or Copenhagen interpretation of quantum mechanics). For example, a "particle" is not assigned a position until its position is measured. In this interpretation, the question of cause and effect is avoided, because quantum mechanics is simply a tool to predict the probability of what is observed.

In the double-slit experiment when a detector was placed before the two slits, a 2 strip pattern was produced after the two slits. When there was no detector placed before the two slits, a different pattern was produced after the two slits. Why does the presence of a detector before the two slits cause a different pattern to be produced after the two slits?
Is there any effort to mitigate the effects sensors have on the quantum particles being observed?
In quantum mechanics, we cannot get information about a system without disturbing it.

#### atyy

In the Feynman lecture (which you really should watch), he sums it up as follows (in the case where thjere is no detector):

Proposition: the electron must either pass through slit A or pass through slit B.

He then shows that, using the experimental results of the double slit, that proposition fails. That proposition seems so close to basic logic that it is hard to understand how it could fail. But, the proposition fails, nevertheless. You cannot say that the electron passed through either slit A or slit B; and, you cannot say that the electron passed through both slits; and, you cannot say that the electron passed through neither slit. You cannot say the electron behaved like a particle; you cannot say the electron behaved like a wave; and you certainly cannot say that an electron behaved like a wave until it was observed and thereafter it behaved like a particle.
Feynman was wrong on this point. It is conceptually possible that the electron went through either A or B. https://en.wikipedia.org/wiki/De_Broglie–Bohm_theory

#### PeroK

Homework Helper
Gold Member
2018 Award
Feynman was wrong on this point. It is conceptually possible that the electron went through either A or B. https://en.wikipedia.org/wiki/De_Broglie–Bohm_theory
Okay, but Bohmian mechanics generally undermines most of what can be said about orthodox QM.

And, invoking Bohmian mechanics hardly helps someone understand the basics of QM.

#### DrChinese

Gold Member
In quantum mechanics, we cannot get information about a system without disturbing it.
The below reference could be considered a counter-example to your reference. Interference is made to disappear without detection (and without any discernible disturbance, other than change of context). And even if (which slit) information were to be obtained, the act of detection is not responsible for the change in outcomes. I would challenge the idea that the system was "disturbed" as part of obtaining information. Only the relative orientation of the polarizers changes - nothing else.

#### ray3400

I now understand how and why observation in quantum mechanics is handled differently than in everyday life, or classical Newtonian physics. Thanks everyone for your responses.

#### atyy

The below reference could be considered a counter-example to your reference. Interference is made to disappear without detection (and without any discernible disturbance, other than change of context). And even if (which slit) information were to be obtained, the act of detection is not responsible for the change in outcomes. I would challenge the idea that the system was "disturbed" as part of obtaining information. Only the relative orientation of the polarizers changes - nothing else.

No it does not challenge it. The author means it in a precise sense, and gives it as Theorem 2.

#### atyy

Okay, but Bohmian mechanics generally undermines most of what can be said about orthodox QM.

And, invoking Bohmian mechanics hardly helps someone understand the basics of QM.
I don't think Bohmian mechanics undermines most of what can be said about orthodox QM. I agree it is not helpful to invoke BM, but unfortunately, Feynman brought up the topic and made an error. I think it undermines orthodox QM, if errors about orthodox QM are left uncorrected.

#### ray3400

Yes, see nondemolition measurements. But this doesn't apply for the double slit.

The pattern changes since a detector at the slit has a serious impact on the microscopic stuff going through the slit:

The change in an observed object due to an observation is large if the means of observing it is comparable (or larger) in size and impact to the observed object. If you magnify the situation sufficiently strongly, it is qualitatively like (though not really like) observing a sand castle by a big wave from the shore.
Thanks, this makes a lot more sense. A lot of YouTube videos make it seem like there is some sort of magical thing that happens - "the particle is aware that it is being detected and therefore behaves differently". As though the particle has some sort of consciousnesses and the change isn't just a result of cause and effect relationship between the measurement device and the quantum particle.

#### vanhees71

Gold Member
Indeed, that's true, as my discussion in #18 showed. You don't need to put additional detector in front of one or both slits to gain which-way information. It's also sufficient to place your detector, i.e., the screen where the particles are registered close enough to the slits to be sure through which slit each particle came, but then you cannot observe the interference pattern as detailed in this posting too.

On the other hand, also the gain of which-information through detection at the slit generally destroy the interference pattern. How to explain this depends on the specific way of the detection.

One simple example is to use quarter-wave plates, oriented in a 90-degree angle relative to each other in the double-slit experiment with polarized photons. If you start with photons polarized at an angle of $0^{\circ}$ in the plane perpendicular to their momentum (which is parallel to the plane of the slits) and orient one quarter-wave plate in one slit at $45^{\circ}$ and the one in the other slit at $-45^{\circ}$ you mark the photons themselves such that you can gain which-way information at any distance from the slit, because the any photon coming through slit 1 is left-circular polarized and any photon coming through slit 2 is right-circular polarized and thus the parts of the corresponding electric-field-operators (note that for photons I cannot argue with wave functions as in #18, because photons do not have a position operator nor a well-defined "wave function") annihilate/create photon states that are perpendicular to each other and thus there's no interference term in the detection probability for these photons at a CCD cam placed no matter how far away from the slits and thus no (two-slit) interference pattern.

Now you gained which-way information with 100% certainty and have no interference pattern. If there were no quarter-wave plates you'd get an interference pattern but are left with no which-way information whatsoever. Of course the difference in the two settings are thus these quarter-wave plates and here of course the reason for the disappearance of the interference pattern is the placement of these quarter-wave plates in the slits and the interaction of the photons (i.e., the em. field) with them, leading to the change of the polarization from linearly polarized to either left- and right-circular polarization depending through which slit the photons came. This very setup has entangled the photon momentum with its polarization in such a way that the state of a photon behind the slits allows with 100% certainty to know through which slit it came.

You can in such a setup also choose to know only "somewhat" through which slit each photon came, i.e., you just orient the quarter-wave plates not with $90^{\circ}$ relative orientation to each other. Depending on which angles of the plates you choose, the probability for a photon with a certain polarization state behind the screen is larger to have come through one slit than the other (say 70% through slit 1 and 30% through slit 2). Then it's more likely that the photon came through slit 1 (70% probability) but you cannot be very certain about it because there's still a 30% chance of having come throuh slit 2. In this case you get in turn some interference pattern back, but not of as high a contrast as the one without the quarter-wave plates in the slits. You can also choose, not to get any which-way information with quarter-wave plates at the slits. Then you simply have to orient them in the same direction. Then there's 50% change for the photon to have come from either slit and that's the most uncertain state concerning which-way information the photons can be prepared in. Then you get the two-slit interference pattern with full contrast.

Indeed, as you see you don't need to really detect the photons to gain the which-way information, but an addition like the quarter-wave plates can lead to "imprinting" this information into the photons themselves, but that's so because of the interaction of the photons with the quarter-wave plates.

So there's some objective reason whether there's which-way information, no which-way information, or some uncertain information, which excludes and interference pattern, leads to an interference pattern of maximal possible contrast, or leads to an interference pattern with less contrast in these three cases, respectively.

However according to the laws of QT you cannot have both, i.e., certain which-way information and an interference pattern. These are mutually exclusive possibilities to prepare the photons.

What's also very clear with these examples is that there's no such thing as "wave-particle duality" or any other "weird things" attributed to QT as soon as you accept, as a fundamental law of nature that cannot be explained by any simpler laws, that there's some "irreducible randomness" in nature, which however is precisely described by modern quantum mechanics or, as here for photons, relativistic quantum field theory.

As @DrChinese said above, physics does not explain "why" nature behaves the way she does, she just accurately describes "how" nature behaves. Sometimes you can explain "why" some phenomenon is observed as it is by using theory to "explain" it by calculations according to the fundamental laws described by this theory, but there's no "deepter explanation" of the "why" of these fundamental laws.

Despite some criticism in this forum, I consider the introduction to (non-relativistic!) QM in the Feynman Lectures as among the best there are in the textbook literature. That's because Feynman is a role model for me in explaning QT as it is without any ado with strange philosophical ideas about some apparent paradoxes (there are none within the minimal statistical interpretation). He just sticks to the physics facts and leaves it at this: The probabilistic nature (i.e., Born's rule to intepret the wave function or more generally any quantum state) is just a fact of nature, deduced in about 120 years of experience with quantum theory (or 94 years if you count only the modern quantum theory, which is the one still up to date with today's knowledge).

Last edited:

"Question About The Role of Observation in Quantum Mechanics"

### Physics Forums Values

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving