# Role of observer and detection in uncertainty principle

1. Mar 7, 2005

### bursar

I'm wondering if someone could tell me what would be the result of the following thought experiment, to help answer a simple question I have about the Heisenberg Uncertainty Principle:
(I am just an educated layman trying to figure it all out- hopefully my thought experiment isn't complete nonsense!)
I perform a double slit experiment. The detector consists of some kind of photosensitive plate to detect an interference pattern. In addition I have a row of lights connected to regions on the detector, each of which will blink on whenever a photon lands in that part of the detector.
A light source is used that fires photons one at a time through the slit (or maybe a beam splitter?) onto the detector at a slow enough speed that an observer can actually see the lights blinking on and off.
I employ an observer to watch the lights blink on. The observer is placed on a comfortable couch for several hours to watch the lights. Eventually he or she falls asleep and stops observing it.
My question: Will an interference pattern appear on the detector since the observer is no longer observing?
My current understanding is that it won't, since by merely connecting the lights to the experiment, regardless of whether someone is observing the lights, the probability wave will 'collapse'.
In that case, if the lights are removed from the experiment, aren't the individual atoms in the photosensitive plate sufficient to collapse the interference pattern?
My basic question revolves around what exactly does it take to make the light behave like a particle and not a wave? Popular press articles make this out to be something of a mystery.

2. Mar 7, 2005

### SpaceTiger

Staff Emeritus
I think you've just inadvertantly rephrased the Schroedinger's cat paradox. Without an observer, the answer is unknown, just as we can't know what's inside a black hole, but the theories will still give predictions. I think the popular interpretation is a superposition of states (particle and wave-like behavior).

3. Mar 7, 2005

### tdunc

no no, I think your confused here with regards in how the word 'observation' is used to detect the position(among other properties) of a particle. See in the quantum world to obserb the position of a free particle you need to fire a photon at it and have it send back info, by doing so you physically alter the particles properties- including location. When you "measure it" or "detect it" you have knowingly changed it. In the case of the 2-slit experiment, The photon that hits the detector is changed by the act of hitting the detector correct? - this is your "observation" the "wave function collapse" or "detection"

So to answer your question: Will an interference pattern appear on the detector since the observer is no longer observing? Absolutely. If it didnt that would be insanely unlogical.

Well really, I'm not sure that the Heisenberg Uncertainty Principle has anything to do with the 2-slit experiment.

4. Mar 7, 2005

### SpaceTiger

Staff Emeritus
I just realized that the person was measuring the positions of the particles after they landed, not which slit they went through. If that's the case, then you're right, it's the same old interference pattern no matter what. I'm assuming the person wanted to collapse the wave function, so I'll proceed as if they were watching which slit the photon passed through.

I understand that, but if nobody is watching the detector, then there is nobody receiving the info, just as with Schroedinger's cat. I'm assuming here that the detector is not recording the state of the photons for the person to see when they wake up -- that would be the same as if they were awake.

It all depends on which interpretation you subscribe to, though. Some people say that the state is changed by the interaction with the particle with which you're observing, not the observation itself. Unfortunately, without someone to look at the system, there's no way to answer that question.

It has everything to do with it. By observing the photons passing through the slit, you're isolating their position, dramatically increasing the uncertainty in their momentum (as measured by the wavelength of the interference pattern) and "collapsing" the wave function.

Last edited: Mar 7, 2005
5. Mar 7, 2005

### dextercioby

:surprised There's a nice chapter in the third volume of Feynman lectures covering this very aspect...

Daniel.

6. Mar 7, 2005

### ZapperZ

Staff Emeritus
I would like you to know that there are many of us who do experiments and walk away from our detectors to get coffee, talk to strangers, go to the bathroom, etc. In fact, many crystal diffraction experiments typically performed at synchrotron centers all over the world run for hours (even overnight) without human intervention or observation. It would be horrible if the universe simply depends on human observation to make it so.

I think what you want to consider is what would happen if you have another detector right at one of the slits that you can randomly turn on and off to detect if a photon passes through that slit, or the other. Now THAT would mess up a lot of people's heads! :)

Zz.

7. Mar 7, 2005

### bursar

Thanks for the posts. I probably sacrificed too much clarity for the sake of keeping the original post brief, nevertheless I think my question has been answered one way or another.
My original experiment as I envisaged has two detection mechanisms. The first is a photosensitive plate, maybe some kind of highly sensitive CCD or film. It would be kind of analog in measurement capability and would not be specifically able to tell you what photon arrived where and when but would be capable of recording an interference pattern for future viewing. The second is a set of on-off sensors connected to regions on the plate capable of detecting individual photons. The output from these sensors would be lights or LEDs and wouldn't be connected to any recording device.
SpaceTiger has added some new information which I wasn't aware of before: to collapse the wave you have to measure the photons before they hit the collector. If this is the case I would have to change the experiment so that a downconverter is used (Ive been getting my ideas from a webpage http://www.fortunecity.com/emachines/e11/86/qphil.html) for each path the photon may take to separately measure whether a photon has taken one path or the other. My observer now only has to watch two lights - left and right, only one of which may be on at a time, to know which path the photon took.
It seems pretty clear now that for the new experiment the wave function will collapse even the observer is absent, even though the path of each photon is not being recorded, since we are doing something to the original photons before they strike the collector plate.
On the other hand, since the downconverters each create another photon (which is how we detect the presence of the original photon), could those photons also be used to create an interference pattern and if so how might their detection method affect the original wave function?

8. Mar 7, 2005

### caribou

Wave function collapse and the importance of conscious observers were both ideas introduced to avoid some problems in quantum mechanics... but they create other problems that are worse then they problems they were created to avoid!

Whether or not a particle exhibits wave or particle behaviour can depend on whether or not it interacts with something.

In the two slit experiment, you can arrange it so that the particle travels through a vacuum and then you get evidence of only wave-like behaviour during the particle's journey and an interference pattern builds up... or you can arrange it so that the particle travels through a gas and then you get evidence of only particle-like behaviour during the particle's journey and no interference pattern builds up.

Interactions with a gas, measuring device, etc. can destroy wave-like behaviour without wave function collapse or conscious observers being needed.

Decoherence is the name of the topic.

9. Mar 9, 2005

### tdunc

You dont measure or observe anything in that experiment - prior to it hitting the wall which is technically not a measurement - thats why you dont know which slit it passed through, so your definatly not changing it and therefore has nothing to do with the uncertainty principle because the 'UP' is all about a measurement being taken and the uncertainty of the future position and properties of the particle due to your measurement. So in the case of this experiment when the photon hits the wall(wave function collapse on impact) is the only time an 'observation' is made but then its over, it has no future.

Now the Copenhagen interpretation I know is something more appropiate to relate to this experiment.

Caribou thats interesting about the 2 setups, the one through gas does not create interference patterns...

also

"Interactions with a gas, measuring device, etc. can destroy wave-like behaviour without wave function collapse" hmm

10. Mar 9, 2005

### caribou

There are a photon interactions removing the interference effects mentioned in Feynman's Lectures but the gas one is from a paper by Jim Hartle I read:

http://arxiv.org/abs/gr-qc/9210006

Wave function collapse is a useful idea as long as you don't take it to be something physical as then you end up with faster-than-light collapse effects and also effects without causes when a wave function collapses when nothing interacts with it.

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

Have something to add?