Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

The Observer Effect

  1. Jan 14, 2008 #1
    I keep seeing references to the observer effect in the double -slit experiment where the a
    of observing the photon or electron going through the slits causes a collapse of the wave function. so, instead of getting a cool interference pattern, you get the pattern expected if the light was acting as a particle.

    Has this actually been tested experimentally?

    If so, I have not been able to find it anywhere.

    If anyone can post a citation from a peer-reviewed physics journal where this phenomenon is reported, I would appreciate it.


    Dr. Todd
  2. jcsd
  3. Jan 15, 2008 #2


    User Avatar
    Homework Helper

    How about (among many): Am. J. Phys. 57 117 (1989)
  4. Jan 20, 2008 #3
    My understanding of the article you cite is that Tonomura et al demonstrated the interference pattern resulting from single-electrons being directed one at a time through the double slits....that experiment did NOT involve showing how the pattern reverts to two straight lines when the electrons are observed going through the slits.
  5. Jan 21, 2008 #4


    User Avatar

    Staff: Mentor

    Actually, you get (or are supposed to get) the single-slit diffraction pattern, which still depends on the wavelike behavior of light. More precisely, you get two such patterns, one from each slit, superimposed on each other. But the spacing between the slits is normally much smaller than the "width" of the single-slit pattern from each slit, so the two patterns almost coincide.

    (I don't have any references to experimental evidence at hand. I just wanted to clear up what would be expected.)
  6. Jan 21, 2008 #5
    Okay, so you're saying that due to the observer effect, and the resulting collapse of the wave function, that instead of an interference pattern there results a single-slit diffraction pattern (or more precisely, two that are nearly superimposed).

    I don't doubt that the hypothetical difference between the double-slit interference pattern, and the superimposed single-slit diffraction pattern can be mathematically determined, with utmost precision.

    But diffraction is basically interference, and just to look at them, the double-slit and single-slit patterns are pretty similar, no?

    The reason I'm saying this, is because the result of any double-slit experiment that has actually been done, and that I can find reference to, is visually EXTREMELY rough...well, and then to add the extra experimental burden of illustrating the observer effect (which these experiments were NOT intending to show)...well, I don't think you'd really be able to see any difference.

    In order for this ever NOT to be simply a thought experiment, the imaging technology would have to be vastly improved. But hey, the Tonomura experiment was done almost 20 years ago, surely we've got better capabilities??
  7. Jan 21, 2008 #6


    User Avatar
    Homework Helper

    two straight lines? Not sure what you're saying, but why don't you just go and Google "double slilt experiment" for yourself. you'll find plenty of stuff.
  8. Jan 21, 2008 #7


    User Avatar
    Homework Helper

    ... or better yet, sign up for a lab course and do the experiment yourself.
  9. Jan 21, 2008 #8
    Well olgranpappy, I'd love to conduct that experiment myself... But, not possible for me at this juncture. Time for me to freely admit, I'm not a university physics student.

    It still begs Dr. Todd's original question....this experiment has NOT been done?? I've googled this extensively, and especially have read in detail about the experiment you cited...

    I'm referring to what Dr. Todd mentioned:
    which is what's supposed to happen when the attempt is made to observe the particles going through the slits, because the wave function is collapsed due to this observation.

    Feynman described the way a detector could be placed at each slit, and electrons detected passing through one or the other, and that simply by performing this detection of which slit the electron passes through, the interference pattern disappears, and the pattern is instead what would be expected from particle behavior. This was often presented by Feynman as a **thought** experiment, to describe in a nutshell fundamental quantum paradoxes.

    So, in the experiment conducted by Tonomura at Hitachi Labs [Am. J. Phys. 57 117 (1989)] they simply demonstrated the wave-behavior of electrons fired one at a time through double slits. They did not additionally demonstrate the disappearance of the interference pattern when an attempt is made to determine which slit each electron passes through....i.e., the "observer effect".
  10. Jan 22, 2008 #9
    check the afshar experiment he showed "observer effect" and "many worlds" is misinterpretations (*watching out for haters*) :P
  11. Jan 22, 2008 #10
    Remember though that you can't just place some "detectors" at the slits without COMPLETELY changing the experiment. You can't measure something without changing it.
  12. Jan 22, 2008 #11
    QUOTE=K.J.Healey] Remember though that you can't just place some "detectors" at the slits without COMPLETELY changing the experiment. You can't measure something without changing it. [/QUOTE] Of course you can, that is the whole point of the observer effect (Which is NOT what Afshar was doing in his Which Way effort).
    The problem was found early in the 20th century using electrons, and was fundamental to the whole “Which Way” Issue.

    Using electrons sent one at a time through a double slit they detected dispersion with interference. The idea was to detect (even if it molested) any electrons going through slit B with a beam of photons that trip a detector just like the seeing eye on your garage door opener.
    Sure that would change those electrons so the checking the screen hits when they got B side detections they were not interested in the results of those “molested” electrons. They want to eliminate those results and only consider the remaining Screen hits, of the same number as the B side hits but Known not to go through the B side. By eliminating all B side hits they then knew they only had A side electrons. Even rate of production of these A side unmolested electrons was confirmed the same as the B side screen hits with detection.
    Now they had a collection of A side electrons going through while the B side slit was still open, yet unmolested and unchanged.
    The result – same dispersion pattern, but without the hoped for interference pattern imbedded within it. By obseving nothing, they affected the results.

    This pre-slit detection was the first to display the Which Way problem. Even bad documentaries like “What the Bleep” can describe this.
    All the OP is looking for is a reference to a real experiment of someone producing these results -- who when and where did they do it.
    Doesn’t someone out there know of a paper or text book that credited someone as actually doing the in front of slit detection experiment? Whitout it how could have QM even known of the paradox needing to be solved.
  13. Jan 24, 2008 #12
    you said "bad documentaries like what the bleep"...do you know any better ones?

    back on topic: I think that the observer effect is in the act of measurement itself since measurement cancels the superposition of probability amplitudes and results in determination, that's why we don't get interference with the observer effect in action. Thus it doesn't depend on how it was measured but rather on the measuring itself.

    However, i too don't know of any actual experiment or paper that says so.
  14. Jan 24, 2008 #13
    Try any Particle Physics or Astrophysics program on PBS, Discovery or The History Channel, - String Theory included.

    As to “back to the topic” - the topic was not questioning the philosophies or various interpretations of what the observer effect might mean. Or even documentation of “entangled” delayed choice types of observer effect observations.
    The OP was very specific; where is any documentation of real experiments using observations of elections as they go though a double slit on the way to a dispersion or interference pattern being observed on a detection screen.
    Preferably, who was among the first to actually correlate screen hits with the failure to detect them going though with other electrons using one slit, therefore confirming they had to have gone though the other slit unobserved. Who where when and how was this first experimental test of knowing which electrons went through one of two slits without directly observing them, and that gaining this knowledge was enough to destroy the ability to find the inference pattern for those electrons.

    An important step to idea that the knowledge of the path taken was as good as a direct observation of that path being taken is enough to prevent interferance due to the "observer effect". You would think that such a clever experiment would warrant a text book or paper somewhere giving credit to someone for it.
    We know when and how Faraday did the first experiments on magnetism, why is this piece of history so hard to find – at least I’ve been unable to find it.
  15. Jan 24, 2008 #14


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor


    1.. X.Y. Zou, et al. Phys. Rev. Lett. v.67, p.318 (1991).
    2. E. Buks, et al., Nature v.391, p.871 (1998).

    There are plenty more of these "which-way" interferometers experiment.

  16. Jan 24, 2008 #15
    Nope; these are not what the OP was looking for.
    Your just skimming the thread without reading for content.

    1.. X.Y. Zou, et al. Phys. Rev. Lett. v.67, p.318 (1991).

    "Induced coherence and indistinguishability in optical interference"
    X. Y. Zou, L. J. Wang, and L. Mandel

    “Second-order interference is observed in the superposition of signal photons from two coherently pumped parametric down-converters, ……….”

    2. E. Buks, et al., Nature v.391, p.871 (1998).
    "Dephasing in electron interference by a 'which-path' detector"
    “ …..Such a manifestation of the complementarity principle was demonstrated recently using a pair of correlated photons, with measurement of one photon being used to determine the path taken by the other and so prevent …….”

    These are “Two Correlated Photons” experiments that are all the rage for some time now.

    The request was for Single photon at a time with direct observation at one of the slits (and therefore no observation at the other slit) type experiment, that dividing hits on the screen into two groups indicating which way for each group.

    I’ve only heard this represented as being done with electrons not photons, but I’ve never seen a reference to who actually did it, either first or as a confirmation.
  17. Jan 24, 2008 #16


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    No, they ARE what the OP is looking for.

    The problem here is in how one actually determine when a photon passes through one of the slit, or maintain the superposition, without destroying the photon. You can't simply put a photodetector at one of the slit and expect that photon to pass through unimpeded. That is why they used a correlated photon as a "twin". By acting on the twin, and by knowing how they are correlated, one has essentially made a determination when it is passing through a slit, or both slit.

    So this is exactly the double slit experiment as requested. Read the paper, not just the abstract.

  18. Jan 25, 2008 #17
    Not if you actually read the OP;
    “where the of observing the photon or electron going through the slits causes a collapse of the wave function”

    Not figuring it out by using one of a pair of photons or electrons as a proxy for detecting the other going through a certain slit.
    Of course that is nearly impossible to do, that’s why posts 11, 13, 15 referred to the electron version of the experiment as the one claimed by some texts as the one actually being preformed. I recall reading them but don’t recall the books or if they gave credit to anyone doing the actual test.

    I’ll draw the picture for you again – I would have though you knew this kind of stuff.
    The main point of the experimental proof was not the directly observed electrons but the indirectly observed electrons. As already described you can detect single electrons passing close to just one slit that they have a chance of going through; without detecting electrons close enough to the other slit that they might go through. Using those detections they could match them with individual hits on the pattern detection screen. By detecting half of the screen hits as coming from the tested slit they could know without directly testing of interfering with the other half that they had to have gone through the other slit. Now they had three definable groups of Electron hits to examine for the pattern produced by each the full group and two half groups one path detected the other not directly path detected but still of know slit origin.
    The results are expected by us now, but not by them at the time.
    When the one slit detection equipment was turned on the full group pattern changed from interference to just dispersion
    And of course the detected group pattern also showed no interference, after all those electrons were getting randomly beat up by a photon beam designed to detect them as they entered or exited the slit on one side.
    Obviously, the pattern being created by the other half of the electrons was being masked the detected half in the whole larger group – right.
    But, no when they examined the pattern of the remaining group of untested at the slit electrons that group also showed just a dispersion pattern.

    The point as least implied in the text books is that these types of experiments were the first to indicate that knowledge even obtained indirectly was enough to disturb the wave like character of individual particles.
    AND that dealing with this information about how realty worked was an important part of defining HUP and Copenhagen. (Unless Zz you have some reference that they were doing correlated "twin". experiments prior to 1926)

    So the question remains – does someone or some several merit credit for actually providing real experimental results that influenced Bohr & Heisenberg. Or were these just thought experiments that were never actually preformed?

    I believe they were done, but have not been able to find a record or resource to confirm it.
    Well I’m going to trust the author when his abstract says they are not using a single particle (one at a time) test.
    If you think their abstract is a misrepresentation, provide an excerpt from the full text. I’m not going to subscribe to a service or pay for a paper I don’t need.
  19. Jan 25, 2008 #18


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    Fine, let's look at the OP again:

    It is asking for the situation where if I don't observe which way the electron or photon goes, I get the interference pattern. If I observe which way the electron or photon goes through, I observe no interference pattern.

    Now, do we agree on this?

    If we do, then I can argue that there are several ways to skin this cat, and not necessarily forcing it to go through a physical slit! All I need to do is set up an experiment, such as a which-way experiment in an interferometer where I give a photon the same 2 distinct path to go through, but I can choose either to know which path it takes, or don't know which path it takes. The key here is to test the the superposition of path principle, isn't it?

    Now, if I use a photon, this will be difficult, because once I make a determination of which which it goes, I've killed it in most cases. So how will I know what kind of a pattern it will make? I create 2 correlated photon. I let one goes through the interferometer, while I take the other, bring it to alpha centauri (or somewhere that isn't the interferometer) and then I have the ability to test it (or not). This "twin" of the one that went through the interferometer can give me that which-way information. It is nothing more than a clever way to be able to measure the photon that went through the interferometer without disturbing it! However, you are hung up on the fact that there's 2 correlated photon being created and closed the door on it beyond that, without realizing WHY they are used in the first place.

    While the premise may LOOK easy, the testing of it isn't! That's why in the simplest conceptual experiment, it takes a lot of effort to get a clean result. That's why experimenters have to jump through experimental gymnastics like this. One just have to take a look at the various Schrodinger Cat-type experiments. Just because the Delft/Stony Brook experiment used the superposition of the supercurrents instead of a "cat alive" and "cat dead" states, doesn't mean that they are not testing the IDENTICAL principle!

  20. Jan 26, 2008 #19
    But I don't think you can have interference and that information at the same time, however, isnt'it?
    Last edited: Jan 26, 2008
  21. Jan 26, 2008 #20


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    What information?

    If you look at the twin to know the which-way path, then no. That's the whole point of having that twin.

  22. Jan 26, 2008 #21
  23. Jan 28, 2008 #22
    No we I don’t agree,
    You assume the OP doesn’t care about how you skin the cat.
    Of course there are easy to find references two photon tests taking advantage twin “entanglement” all over the place with a Google search. Which accounts for your terse reply in post in post 14
    “There are plenty more of these "which-way" interferometers experiment.”
    Because yes any idiot should be able to find plenty of examples using paired photons – they are the most popular modern QM experiment. But you only offer two hard to find references without titles that require subscription or payment – was that some kind of dig at the OP or do you just assume everyone has subscription access.

    I instead assume the OP is not an idiot and helped focus the question on the single particle testing (beam of electrons) into a double slit I believe was the intent. Your version has been answered, the single particle version which I detailed in earlier posts has not.
    And I completely agree with the OP finding a reference for who, when, and how the single electron experiment was done is not easy to find.
    I have not been able to find it anywhere either.

    Correlated photon tests were created or discovered with the help of Copenhagen not the other way around. But the much earlier version of the observer effect in the single electron tests was supposedly some of the real experimental data available to Bohr & Heisenberg and therefore was part of establishing Copenhagen.

    Establishing a known group of electrons without directly observing those electrons and long before the idea of “Entangled Particles” was available was no easy task. And IMO much more clever and difficult at the time than creating two particles in entanglement which already had over 30 years a working QM theory to support that “clever” approach.
    This single electron test established the concept of “knowing” without “seeing” the electrons go through a certain slit was enough to prevent interference showing up in the observed pattern.
    You do agree the concept of observer knowledge without a direct observation was an important part of the creation of the Copenhagen Interpretation.
    Aren’t you a little curious yourself as to where they got the real world data to build CI.
    Were they running their own experiments or were they explaining the results of others?
    I do not believe CI was developed from just pure thought experiments, but I cannot find a reference for this type for test being done or repeated before or after CI was put forward.

    If you don’t know, or cannot find it for us just say so
    Maybe someone else can.
  24. Jan 29, 2008 #23
    Never got this. How can the act of observing something change the result? How does it know it's being observed...? ;)
  25. Jan 30, 2008 #24


    User Avatar

    If you try to think that "an observation" is the same thing as "an interaction", your question amounts to asking, why can't I interact with something without perturbing it.

    IMO, somehow the notion of an observation or interaction is to see how something responds to a perturbation. Your question is the perturbation, and the response is the answer.

    So why does questions change the answers? Maybe because the perturbed system is consntantly responding and adjusting, and it's response pattern may change if it's repeatedly are exposed to the same questions. It's similarly clear thta the choice of questions, affects the set of possible answers. The questions themselves are not innocent.

  26. Jan 30, 2008 #25
    Thanks for the reply. I can understand how an interaction would most certainly interfere with the experiment but why does an observation have to be an interaction and not purely an observation. It implies that the experiment somehow knows that it is being observed...
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook