Talk:Double-slit experiment/Archive 6

Archive 1

←

Archive 4

Archive 5

Archive 6

Archive 7

Archive 8

→

Archive 10

Max Born and Triple Slit

I put the results of a recent Science article testing the triple slit hypothesis of Max Born, which states that the pattern will not change when the light passes through three or more slits. This is an important result of the wave/particle duality, because surely if light is waves, then there will be a difference. But I am not sure I put it in the the correct place in the article. I hope some kind soul will check it. Nick Beeson (talk) 21:08, 17 August 2010 (UTC)

Unfortunately the article by Born is not in the public domain.WMdeMuynck (talk) 10:13, 18 August 2010 (UTC)

It would be my advice not to burden the Double-slit experiment article with the results of the paper by Sinha et al. because it does not add anything to our understanding of the Born rule. The experiment performed by Sinha et al. is interesting because it is studying a subject that potentially is of practical importance in quantum cryptography and quantum computation, being of a fundamentally new type not mathematically tractable by means of standard quantum mechanics. However, the result of the experiment turns out not to deviate significantly from the predictions of standard quantum mechanics. Hence, it does not learn us how to make use of the resources provided by generalized quantum mechanics (dealing with positive operator-valued measures).WMdeMuynck (talk) 20:59, 19 August 2010 (UTC)

I agree that the Triple slit experiment is a bit useless -- beyond adding hype to an already over-hyped paper in the journal Science. —Preceding unsigned comment added by 149.157.192.251 (talk) 12:12, 21 September 2010 (UTC)

minor typo fix needed

I have very little time here, not enough to format this comment properly, so I'll just leave you a quick note. There's a very easy to fix typo: In your double slit pictures, from a red laser, the caption lists the wavelength as "0.7 mm".[e.j., millimeters] What you want is on the order of 660 nano-meters (700 would work). I suspect this is simply a typo -- someone wrote "mm" when they meant "nm". (mm is 10^-3; nm is 10^-9.)

That isn't the wavelength; it's the distance between the slits, as clearly stated in the caption. -Jordgette (talk) 01:42, 12 September 2010 (UTC)

Affirming the consequent...

This article is in many ways unusually complete, however, re the section beginning:

"One of the most striking consequences of the new science is that it is not in agreement with the belief of Laplace..."

Firstly, this is terrible weasel-wording. If it is meant that the new science contradicts Laplace, then say that. But since there is no logical contradiction (without affirming the consequent), I don't recommend that course.

The results were consistent with those opposed to Laplace, true; but they don't contradict Laplace, etc. These results may demand some additional explanation from him, some complication for any alternative theory. They may not be what that side of the argument hoped to see. But only by deciding in advance that no amount of creativity can rescue Laplace's idea (no black crow) can you toss it. Putting a heavy spin on one's words (pardon the pun) doesn't suffice.

Historically, scientists have often mistaken consistency of experimental results with the ideas that led them to the experiment with a logical contradiction of all other ideas. But this affirms the consequent (Not to mention "Post hoc ergo prompter hoc".) It would be nice to spare ourselves the necessity of investigation by such quick logic, but surely moves that don't conform to logical principles shouldn't be enshrined as science.

Probability is generally a measure of ignorance. It may be that in quantum matters it is instead fundamental. But formal logic can't prove that physical fact, nor could such a thing be discovered (since this too is "a no black crow" statement.) Of course, evidence consistent with this conclusion can be amassed (which is also consistent with many other explanations, some of which may not be very attractive), and one such example of evidence is cited in this article - and then, unfortunately, conflated with proof, and a proof of something that by its nature, can't be proven.

One instance of "absence of evidence" (of order in the apparent randomness) isn't even strong "evidence of absence."

To forestall another objection: note that only classical hidden variables are ruled out by Bell's entanglement experiment, not all determinate theories of any kind no matter how bizarre.

There are many experiments consistent with indeterminacy, it's bizarre to pick out this one to confuse with a logical contradiction of determinacy. —Preceding unsigned comment added by 24.69.150.0 (talk) 01:28, 16 May 2009 (UTC)

At a minimum this largely extraneous section is false as stated. Move it to the Uncertainty Principle or Copenhagen Interpretation articles (since it tries to preempt those discussions) or maybe just dump it. —Preceding unsigned comment added by 24.69.150.0 (talk) 01:23, 16 May 2009 (UTC)

The following sentence in the article I found misleading: "However, an experiment performed in 1987[9] produced results that demonstrated 'which-path' information could be obtained without destroying the possibility of interference." The fact is, any type of measurement - even the experiment cited effects the consequent interference pattern in some manner. There has been no experiment to date that measures the 'which-path" and still leads to the full interference pattern without measurement. It is sufficiently misleading that most Wikipedia readers will jump to the conclusion that one can measure without consequent effects. —Preceding unsigned comment added by 98.155.237.197 (talk) 04:50, 8 February 2011 (UTC)

Major removal

Somebody removed a large section of this article. I have reverted the change. If people want to change anything of this magnitude then they should discuss the matter first. P0M (talk) 19:49, 19 January 2011 (UTC)

The passage has been removed again, with a comment that it opposes information in the lead.P0M (talk) 00:45, 20 January 2011 (UTC)

The passage has again been restored by me (WMdeMuynck) because, when there is a contradiction (which there certainly is), it is arbitrary to remove a remark based on recent insights obtained from well-cited experimentation in favour of guesses based on 80 years old very restricted experimental evidence. If it is impossible to give an in-depth discussion of the subject here (which I have experienced to be the case), then it seems to me more appropriate for an encyclopedia to put different (possibly contradicting) views next to each other.WMdeMuynck (talk) 18:43, 20 January 2011 (UTC)

Trying to get the discussion of the double-slit experiment up-to-date

The following sentence in the article I found misleading: "However, an experiment performed in 1987[9] produced results that demonstrated 'which-path' information could be obtained without destroying the possibility of interference." The fact is, any type of measurement - even the experiment cited effects the consequent interference pattern in some manner. There has been no experiment to date that measures the 'which-path" and still leads to the full interference pattern without measurement. It is sufficiently misleading that Wikipedia readers will jump to the conclusion that one can measure without consequent effects. —Preceding unsigned comment added by 98.155.237.197 (talk) 04:50, 8 February 2011 (UTC)

I have corrected a paragraph on wave-particle complementarity presenting notions introduced during the days quantum mechanics was based on experimental data obtained by means of a very restricted set of experiments. In particular, in optical experiments it was not possible then to detect impacts of individual photons (interference experiments with electrons stem only from the late 1950s and 1960s). Nowadays we have ample evidence that these impacts are particle-like, thus making obsolete the idea that no particle-like phenomena can be observed in an interference experiment. A related generalization is the extension of the domain of application of quantum mechanics to measurements described by positive operator-valued measures rather than by the hermitian operators of the standard formalism, thus allowing to study the whole range of parameters in a continuous way connecting a pure interference experiment with a which-way experiment.

It is not clear to me that anyone has made the claim that no particle-like phenomena can be observed in an interference experiment. Maybe that is because such a claim is patently incorrect and my poor mind has simply edited-out or reinterpreted such claims. I interpret the statement you removed as indicating that in order to explain the resultant interference pattern it is necessary to use a wave interpretation (that would be "time one") and then make a conceptually awkward switch (covered by fudge factors like "psi function collapse") to a particle explanation to account for the appearance of a photon, electron, or whatever as a spot on the detection screen (at "time two"). P0M (talk)

The idea of particle-wave complementarity can be found in Bohr's contribution to P.A. Schilpp, ed., Albert Einstein: Philosopher--Scientist, The Library of Living Philosophers, 1949, Vol. I, p. 199-241, in particular p. 217/8.WMdeMuynck (talk) 20:40, 31 March 2009 (UTC)

Strange, I had seen the problem, but it never occurred to me that Bohr or Heisenberg might have imagined each photon to have "deposited" itself with varying intensity across all of the fringes of an interference pattern. How would they have accounted for the photoelectric effect then? As resulting from each photon contributing to a disturbance across the entire surface of the photoelectric cell? I have found before, in other fields, that it is hard to avoid anachronisms. but probably I have just learned this lesson one more time. P0M (talk) 06:41, 1 April 2009 (UTC)

For the non-specialist reader like me it would be helpful if you were to explain expressions such as "positive operator-valued measures", "hermitian operators of the standard formalism", etc.P0M (talk)

You might consult my website, to which a link is provided on my User page.WMdeMuynck (talk) 20:40, 31 March 2009 (UTC)

I realize that, maybe, my correction is a bit precipitous, and should better be introduced in a more elaborate way. For me doing so is no option, however. On the other hand, I hate reading in Wikipedia the obsolete information physics students probably learn even today from teachers telling them what they themselves learned a long time ago.WMdeMuynck (talk) 15:57, 31 March 2009 (UTC)

It is highly desirable to replace obsolete information. However, to me, the change you have made describes a misinterpretation but does not supply a correction thereof.

It is a widespread misunderstanding, originating with the Copenhagen interpretation, that any modification of the apparatus that can determine which slit a photon passes through destroys the interference pattern, thus illustrating the complementarity principle in the sense of particle-wave complementarity.

To me, this statement implies that there is a modification of the double-slit apparatus that can determine which slit a photon passes through and yet permit that photon to contribute to the formation of an interference pattern. If that is the case, then it would be essential to provide citations to the relevant experiment. P0M (talk) 17:07, 31 March 2009 (UTC)

My reference to Mittelstaedt, Prieur and Schieder presents the first experiment showing that the interference pattern is not completely wiped out when the experimental arrangement is modified so as to yield path information. Many other examples are discussed on my website or can be found in the papers referred to in the main publications list on that site.WMdeMuynck (talk) 20:40, 31 March 2009 (UTC)

Are you affirming that which-path information can be obtained for single photons, electrons, etc., and yet they will contribute to an interference pattern? P0M (talk) 02:26, 1 April 2009 (UTC)

I concur, This is really vital information. Given that there is contradictory information abound on the issue, any such statements whether or not an observer determining and observing which slit the particle has passed through collapses the wave function needs to be backed up. Personally, I would like evidence from the electron experiment, as I can envisage relatively easily detecting electrons (in comparison to photons), and electrons are considered matter, which makes the experiment even more interesting. An observer collapsing the wave function of matter makes the universe very queer, and if it is not the case, debunking the myth is a general good. Nick R Hill (talk) —Preceding undated comment added 01:04, 24 May 2009 (UTC).

There is no serious doubt that "an observer determining and observing which slit the particle has passed through" can destroy the possibility of its taking part in an interference pattern. Heisenberg and the people in his circle avoided the kind of almost solipsistic point of view that says that human consciousness has to be involved. The fact is that some things that can be done to identify whether a photon went through slit one or slit two will not only indicate that a photon was at one or the other of the slits, but it will also spoil that photon's contribution to any interference pattern, i.e., it will not interfere with itself. The conclusion of many people has been that anything that identifies which slit a photon went through will inevitably prevent that photon from interfering with itself. Their argument is that the photon has to be "not at" either slit to interfere with itself, and that if it is "at" one slit or the other it is "all at" one place or the other and therefore cannot interfere with itself. Feynman argues that if one locates a particle at slit one or slit two then the path of the particle is from the laser (or other source) to the point at one of the slits, and that anything that shows up at the detection screen is another event involving a successor photon whose path is from that slit to the detection screen and therefore does not involve the other slit.

The counter-argument is that it is possible to so delicately interact with a photon at one or the other slit that one can detect its presence without totally destroying the interference phenomenon. (See the discussion below.) So far nobody has been able to avoid the appearance that some of the particles are detected at one or the other slit, and do not participate in interference, and others are not detected at either slit and do participate in interference.

However you cut it, the basic fact seems to be that there must be two (or more) slits, and that "something" has to go through both slits in such a way that the two "somethings" can come together and so interfere with each other. Beyond that, people seem to have the most difficulty with the idea that something with mass, such as an atom or even a molecule, can be part here and part there. They believe that the mass either goes through one slit or the other, and that "something else" goes through both slits, interferes with itself, and thereby accounts for where the mass will show up on a detection screen. I think that nobody seriously doubts the idea that one can do "enough" to a particle in the vicinity of one slit or the other to prevent its forming part of an interference pattern. Then I guess you would have to ask what doing something to the electron in one place nullified the function of "that other stuff" that was going through the other slit at about that time.P0M (talk) 05:29, 24 May 2009 (UTC)

In a way the "which-path" information is utterly simple and trivial. you close one of the slits. whether you physically do this, elecromagnetically do it, or what have you. any operation that enables you to count only those photons that go through the area of space called "slit a" unabated, when you add those counts together, will give you an interference pattern equal to that which you would see if slit a was the only slit. it's a simple matter of how counting works. you simple did not count the other paths, for whatever reason.Kevin Baas^talk 15:43, 21 February 2011 (UTC)

Huh? If you close one of two slits there is no longer any interference pattern. There is only a diffraction pattern. The appearance of the detection screen when there is only one slit open is very different from the appearance of the detection screen when there are two slits open (and they are of the appropriate size and distance from each other).P0M (talk) 01:28, 22 February 2011 (UTC)

Also, it is not necessary to stop anything from getting through the apparatus to show up at the detection screen in order to frustrate the formation of an interference pattern. Extending the narrow strip between the two slits in the direction of the detection screen and then vertically so that there is a wall between whatever travels through slit A and whatever travels through slit B will prevent interference. Also, putting a polarizer in vertical orientation directly in front of one slit and a second polarizer in horizontal orientation directly in front of the other slit will also prevent interference. Except for the occasional photon that gets picked off by an electron in one of the polarizers, all the photons will make it through to the detection screen. Nothing is "closed off." Moreover, it is not some sort of solipsistic "human consciousness interaction with the Universe" or anything like that. You could run the experiment with a laser that produces only one photon at a time, and you could count the arrival of photons at the detection screen, and you would get an interference pattern building up without the polarizers in place, and you would not get an interference pattern building up with the polarizers in place. The reason is that the superimposed states of what arrives at the detection screen in each instance with polarizers are superimposed states that have oscillations that are 90° out of phase. P0M (talk) 14:36, 22 February 2011 (UTC)

So in these examples you have given you are stopping photons from getting through the apparatus in the approximate position of the second slit. How is that not doing precisely that? Kevin Baas^talk 15:12, 22 February 2011 (UTC)

Not so. If there were a theater with a door for women on the left and a door for men on the right, and there were a barrier right down the middle so that they could never get together as they moved from the door they came in to a chair somewhere up on the stage, you would prevent any men from showing up with the women, and you would prevent any woman from showing up with the men, but you would not stop any of them. Being truly luminous beings, they would not loiter along the way but each would take a seat on her or his side of the partition.

Similarly, if you wanted to prevent men and women from performing the Platonic reunion on stage, you could cause all the women to have their formerly wild gyrations constrained so that they gyrated only in a diagonal mode, and all the men to have their formerly wild gyrations constrained so that they gyrated only in the anti-diagonal mode (or one could be horizontal and the other vertical, same difference). You would not stop any of them from arriving at the distant screen, but when they got there they would be effectively prevented from any disgracing interactions.

You realize that by "wall" i am not demanding a perfectly literal physical wall made of brick and concrete, but i am speaking figuratively? Kevin Baas^talk 14:30, 23 February 2011 (UTC)

I was the one who wrote about a wall, not you. I could have spoken of a partition without changing my meaning. I don't see where you have spoken about a wall yourself (up to this point I mean) -- figurative or not. P0M (talk) 05:31, 24 February 2011 (UTC)

Eh, perhaps not. in that case i am refering to "close one of the slits", or in any case an apparent difference in the meaning we apply to the term as it applies on a quantumn scale. (as i've discussed below.) Kevin Baas^talk 13:59, 24 February 2011 (UTC)

Also above you state that feymann says that by observing the particle at one slit, well, you say feyman describe this as stopping a photon and starting a new one. that is, you say that in feymann's description, as you characterize it, it is absolutely neccessary that something is stopped. how do you rectify this with your statement that it is not neccessary to stop anything? that seems to me like a contradiction, unless you are saying that feymann's path integral formulation is wrong. which would be a bit presumptuous, imo, as i am not aware of any experiments that don't agree with it perfectly. Kevin Baas^talk 19:36, 22 February 2011 (UTC)

What Feynman says is not really exclusive to Feynman. He just has a nice way of saying it in one of his books. What happens in the basic experiments that got quantum mechanics started (because the experiments were showing them things they could not account for according to classical physics) is that a photon come into the physicist's apparatus from somewhere, it is absorbed by an electron in its orbit around the nucleus of some atom, and after a while it drops back to its equilibrium position, and in so doing a photon is emitted. That is the vanilla situation that you would find in examining the absorption and emission characteristics of a tube full of helium gas. So what Feynman said, was that if a photon enters a physics apparatus and we think we would like to know whether it went from S (the source of the light) through slit A or slit B and then arrived at some position on the detector screen D, we may put in detectors A* and B*. Starting on the bottom of p. 81 of QED, he says, "To calculate the amplitude that the detectors at A and D go off, we multiply the arrows that represent the following steps: a photon goes from S to A, the photon goes from A to D, and the detector at D goes off.The other complete event is the detectors at B and D go off." You will note that he speaks of the same photon as going from A to D. I think the difference between his way of speaking and mine is verbal only. The crucial things is that the photon either shows up at A or it shows up at B. It it "shows up," if it is measured, then it has to have come into interaction with the lab apparatus. The only way that a photon can come into interaction with a lab apparatus and leave something observable is to contribute some energy to the detector. To do so, the photon has to be absorbed in boosting an electron up to a higher orbital. (Study of the energy changes and the consequence that the photons emitted by such a detector at a very short time later will change in frequency and direction was the very experiment that Heidegger worked on with Kramers that gave him a flying start at the new quantum mechanics. (H. A. Kramers and W. Heisenberg, “Uber die Streuung von Strahlen durch Atome”, Z. Phys. 31, 681-707 (1925))

If a photon gets stopped, or, to use another description, if the terminals of an event are that it starts at a laser and is detected at x,y,then what goes on from there is a whole new ball game. It's not "one event" by Feynman's way of speaking. It's not "one photon" in my way of speaking, and of course I would not count things as "one event" either.

All I said before was that with just the two slits there nothing gets stopped. If you detect something, you stop something. If you do not detect anything, you do not stop anything.

There is another alternative, as indicated above. You can use polarizers. That way you could know whether a photon had gone through slit A or through slit B by putting another polarizer in one or the other orientation before the detection screen. If a photon did not make it through, you could conclude that it was from the slit that had the other orientation. If a photon did make it through, you could conclude that it came by way of the slit that had the same polarizer orientation as that third polarizer. However, when you do that then you no longer get an interference phenomenon. So even constraining the photon enough to give it a definite/defined polarity will destroy the possibility of its interfering with itself and contributing to an interference pattern.

So far we have gone back over ways of definitely stopping a photon, and then letting "it" go on again, and ways of sort of branding it so that we could tell by later measurement which slit is consistent in its polarity with what we get near the detection screen. However, there is another way of preventing interference from happening. That is to build a partition. Now this is a really weird idea, at least if you are one of those who believes that the photon really either goes through slit A or slit B. There is nothing interfering with the photon if it goes through slit A, and there is nothing interfering with it if it goes through slit B. Supposing somebody worries that the partition exerts a baleful influence somehow, we could put in an angled mirror just beyond each of the slits so that the beam coming through each slit would never impinge on the partition. If, on the other hand, you believe that something goes through slit A and something goes through slit B, then the partition obviously ruins interference between the two by keeping them apart.P0M (talk) 03:06, 23 February 2011 (UTC)

I already understand all the physics. But now I understand that this is just a difference in our semantics. By "wall" i mean on a quantumn scale, not a classical scale. And thus any thing that significantly changes the or confuses the action along different paths, such as e.g. a polarizer. Or a whole bunch of fairly randomly oriented atoms in a solid state phase, which can hardly be said to be a "wall" as there are vast distances between any two "physical objects" (e.g. electrons, protons, etc.), but succeed in very thoroughly scattering and mixing the phases of any incoming light such that they effect cancel each other out on the other side, and not necessarily by "stopping" anything, in the classical sense. But if we are going to use the word "stop" in this context, or "wall", for that matter, I would think we should find some kind of figurative analog. Thus, by "stop" in the quantum context, i don't necessarily mean that anything "touches" anything at all, or even ceases to move, but simply that after a certain points, some of the actions along paths are such that the some (not necessarily all!) of the phases angles are statistically complementary and thus cancel out (save some insignificant amount). So it seems to me we are simply using different semantics. Kevin Baas^talk 14:53, 23 February 2011 (UTC)

Buckminsterfullerene

I visit this page off and on. When I read Buckminsterfullerene is visible under a microscope, I wanted to know if this was an electron microscope, or a more traditional one? 02:33, 19 February 2011 (UTC)

path-integral formulation - some notes.

i noticed it says "needs expert attention" in that section. i'm no expert, but i did notice that technically it's incorrect, but to make it correct would make it practically unreadable. Particularly the line "In the path integral formulation, the probability distribution of the outcome is the superposition of every possible classical path through space-time to get from point A to point B", apart from being ambiguous, is patently false. The path integral formulation actually says that the probability distribution is the square of the sum (superposition) of all possible paths, where the contribution of each path has unit magnitude, and a phase angle proportional to the classical action of the path.

That is, the contribution of each path is given by: e^{(i/h) * S(x,t)} where h is plank's constant, and S(x,t) is the classical action. Sum that up over all possible paths, square it (i.e. multiply by the complex conjugate), normalize it, and that's the probability distribution; the probability of finding a "particle" at position x, time t.

how this relates to the double slit experiment? well before you square the contributions, some of them will cancel out due to contributions from paths that are opposite in phase, some path contributions will interfere constructively, etc. giving you an interference pattern.

how to write that up, simply, i'm at a loss. Kevin Baas^talk 17:18, 17 February 2011 (UTC)

Wikipedia articles on technical subjects often trend in the direction of increased accuracy but decreased readability. I would rather go in the direction of general accuracy, rather than introduce the math in an effort to achieve perfect accuracy. Achieving perfect accuracy is a fine goal, but if the resulting section becomes offputting or useless to the lay person, then I think it's a change for the worse. In that spirit, may I suggest something like, "the probability distribution of the outcome is related to the superposition of every possible classical path through space-time to get from point A to point B"? If someone wants the math specifics, they should be able to go to the path integral page and get it there. -Jordgette (talk) 23:44, 17 February 2011 (UTC)

better, but "the superposition of a path" is really non-sensical. you can't superpose paths. you superpose some collections of magnitudes over a region. that region might be defined by a path or a collection of paths. but it's not the paths or the regions that you're superposing, it's some function defined over it. it doesn't make any sense to superpose raw regions or paths, not even intuitively.

also "every possible classical path" is misleading, even wrong, as really you must include every path, quite irrespective of whether the path is allowed ("possible") according to classical physics. (in fact there is only one "possible classical path": the one of "least action".)

Simplified, even over-simplified wording is one thing, fundamental misconceptions and just plain nonsensical wording is quite another. I don't mean to be harsh. Just to say that these errors remain and make it both inaccurate in general and manifestly unreadable. And they should be corrected somehow. Kevin Baas^talk 15:04, 18 February 2011 (UTC)

Ultimately, in order to not be nonsensical or inaccurate, one would have to say at least that it is the normalized square of the superposition of [ some function of some classical function ] over all paths. now we can drop the "normalized square of..." part and just say "related to..." instead. the "over all paths" is definitely not droppable, leaving us with only the part in brackets. now it's fairly obvious that the classical function is the "action" as that is the only one with the right units and dimensionality. and it's a lot more clear, without sacrificing readability, to just say "action" (or "classical action"), rather than leaving the reader in some vague and abstract nowhere land of "some classical function", which just adds to the grammatically complexity.

that leaves only "some function of". the function happens to be utterly trivial: e^ix. it called a "phase factor", and in fact all it's saying is that an electron is a wave (see Euler's formula), just like light is. which is a pretty fundamental assertion for the whole theory. hell, without it, it would disagree with the results of the double-slit experiment! Kevin Baas^talk 15:46, 18 February 2011 (UTC)

in fact the whole theory can be summed up this way: "The electron is a wave that has no idea where it is going." a bit comical, but rather illustrative and surprisingly comprehensive. (the entirety of quantum mechanics can probably be deduced from that one statement.) though probably not appropriate wording for the article. :P Kevin Baas^talk 15:53, 18 February 2011 (UTC)

how about something like: "The probability distribution of the outcome is the square of the superposition, over all paths from the point of origin to the final point, of waves propagating proportionally to the action along that path. The differences in the cumulative action along the different paths produces the interference pattern observed by the double-slit experiment." Kevin Baas^talk 21:26, 18 February 2011 (UTC)

I'll wait for more suggestions in re-wording and if there are none after a while i'll go ahead and make the change. Kevin Baas^talk 15:26, 21 February 2011 (UTC)

changed. btw, for sake of simplification, i did introduce a small inaccuracy: it should actually be the square of the norm (absolute value), or otherwise put, the function times its complex conjugate. so that in any case the result is a positive real number valued wave. (w/exactly twice the frequency of the original.) Kevin Baas^talk 15:19, 22 February 2011 (UTC)

I made a couple more changes to this section. Hopefully this is a lot clearer now. Comments welcome. Kevin Baas^talk 16:16, 22 February 2011 (UTC)

eh, now i've introduced an error: it's not exactly an oscillating function of the action, but rather a rotation through the complex number plane proportional to the action. sort of like simple harmonic motion about the origin of the complex number plane, w/phase velocity mediated by the action. hmmm... brings up some curious physical questions, e.g. are we only seeing the stable orbits; are there suborbital and super-orbital rotations about the origin that we can't see in our experiments (or in any case haven't looked for)? etc. anyways, while more accurate, i'm not sure rotation through the complex number plane would be much clearer. :P esp. given that in the end one does get an oscillating function, but then again that's after summing up the rotations. what to do, what to do... Kevin Baas^talk 17:14, 22 February 2011 (UTC)

So... do we still need the "needs expert attention tag" or can that go now? Kevin Baas^talk 14:10, 25 February 2011 (UTC)

A remaining problem.

I don't know whether this matter requires "expert attention" or just some careful thinking about what is being said by our resident physicist, what is said in the article, and how the reader new to the subject will interpret the article as it is currently written.

Some time ago a longer exchange resulted in a sort of summary judgment:

My reference to Mittelstaedt, Prieur and Schieder presents the first experiment showing that the interference pattern is not completely wiped out when the experimental arrangement is modified so as to yield path information. Many other examples are discussed on my website or can be found in the papers referred to in the main publications list on that site.WMdeMuynck (talk) 20:40, 31 March 2009 (UTC)

Are you affirming that which-path information can be obtained for single photons, electrons, etc., and yet they will contribute to an interference pattern? P0M (talk) 02:26, 1 April 2009 (UTC)

Then there was some conversation introduced by others, and finally:

I had another question left over from above: "Are you affirming that which-path information can be obtained for single photons, electrons, etc., and yet they will contribute to an interference pattern?" P0M (talk) 17:22, 9 April 2009 (UTC)

Quantum mechanical information is information on ensembles. For instance, an interference pattern is a property of an ensemble. What has been demonstrated is that by changing the measurement arrangement in a certain way statistical information can be obtained on both interference and path observables. We studied measurement procedures in which a parameter could be changed in a continuous way so as to have the measurement results change continuously from the results of an interference measurement into the results of a which-way measurement.WMdeMuynck (talk) 21:17, 9 April 2009 (UTC)

I think this answer is correct, but that one paragraph near the top of the article gives the wrong impression:

It is a widespread misunderstanding that, when two slits are open but a detector is added to the experiment to determine which slit a photon has passed through, then the interference pattern no longer forms and the experimental apparatus yields two simple patterns, one from each slit, superposed without interference[12][citation needed]. Such a result would be obtained only if the results of two experiments were superposed in which either one or the other slit is closed. However, there are many other methods to determine whether a photon passed through a slit, for instance by placing an atom at the position of each slit and monitoring whether one of these atoms is influenced by a photon passing it. In general, in such experiments, the interference pattern will be changed but not be completely wiped out. Interesting experiments of this latter kind have been performed with photons[9] and with neutrons.[13]

To me, this paragraph seems to attack a straw man, or possibly it reflects an attempt to respond to some very confused reactions by a few readers who have a sort of "magical" attitude toward the effects of observation. To risk making a parody, perhaps some individuals have elevated the idea that a watched pot will not boil to the even more incredible claim that by watching one pot somebody makes all pots fail to boil.

Perhaps some people who have studied the double-slit experiment have come to believe (by whatever thought processes) that measuring a subset of the photons directed at the double slits would prevent the interference of all photons, and that therefore no interference pattern could be found. However, I think it is highly unlikely that such would be the naive or unschooled reaction of most people who learn of the double-slit experiment. I think that their initial (and correct) reaction would be that some causal factor that interacted with a subset of the photons involved in this experiment could only be expected to affect their contribution to what showed up on the detection screen. For instance, momentarily closing one of the two slits with a shutter would only be expected to affect the behavior of photons going through the slit apparatus during that period.

Most people would not expect some action directed at a subset of photons in the experiment to affect the behavior of all the photons involved in the experiment. So if this matter is to be mentioned in the article at all, then I think it ought to be offered as a confirmation of the idea that observing one photon takes away the possibility of that one photon to interfere with itself. The explanatory value of experiments that interfere with a subset of photons but leave the remainder of them alone is that we can see in what amounts to a statistical summation of many individual interactions that this kind of experiment confirms, through what amounts to a huge number of individual "runs" of a single-photon experiment, that photons that have been measured do not contribute to an interference pattern, and that the "fringes" fuzz out in proportion to the amount of messing with individual photons that has been done.P0M (talk) 14:48, 26 February 2011 (UTC)

If nobody disagrees with my reasoning detailed above, then I am going to change the paragraph quoted from the article.P0M (talk) 03:13, 1 March 2011 (UTC)

I think that paragraph is important as it serves to correct some commons misconceptions. Kevin Baas^talk 13:33, 1 March 2011 (UTC)

My advice to User POM would be to refrain from his intention to change the paragraph quoted on the basis of his reasoning detailed above, since this reasoning is irrelevant to the paragraph. Quantum mechanics hardly ever tells anything about individual objects; it is a theory about ensembles. What is important in the quoted paragraph is the influence of the measurement arrangement on the results of measurements on an ensemble. These results, the relative frequencies (probability distributions) change if the measurement arrangement is changed. In early discussions of the two-slit experiment only discontinuous changes were considered (such as between measurement arrangements for measuring Q or P). Nowadays also measurement procedure can be discussed (and carried out experimentally) in which some parameter of the measurement arrangement is changed so as to go in a continuous way from Q to P. I have discussed some examples of these[[1]] and[[2]]. The point is that a continuous change of the measurement arrangement yields a continuous change of the probability distribution.WMdeMuynck (talk) 23:02, 1 March 2011 (UTC)

The problem, as I see it, is that the current article does not reflect what you just wrote, to summarize: "Measurement procedures can be discussed (and carried out experimentally) in which some parameter of the measurement arrangement is changed so as to go in a continuous way from Q to P. A continuous change of the measurement arrangement yields a continuous change of the probability distribution." Probably the mention of Q and P would have to be generalized because the reader would not follow the discussion at this point for lack of context.P0M (talk) 02:46, 2 March 2011 (UTC)

I agree that my reference to Q and P should be generalized (the measurements I referred to are not related in any way to these observables). There is one measurement in quantum optics, called eight-port optical homodyning, that is a nonideal joint measurement of observables Q and P. However, most of the time other observables are involved. It is also true that these generalized measurements are not well-known, and would have to be introduced in a rather extensive manner. I am not going to do this myself because I have done this on my web site already in an extensive way. As you can see from that web site, the problem of generalized measurements is touching upon many aspects and problems of quantum mechanics that are interrelated and cannot be dealt with separately. I am a bit pessimistic about the possibility of convincing the average reader, who has been raised believing that undergraduate quantum mechanics is the ultimate truth, that a more generalized view makes quantum mechanics much more physical and also far less spectacular than is suggested by Schroedinger's cat and Bell's inequalities.WMdeMuynck (talk) 16:31, 2 March 2011 (UTC)

I assume you mentioned P and Q in what your wrote because of the context in which those comments appeared. How about listing some other pairs... "changed so as to go in a continuous way from Q to P, kinetic energy T and position x, or [other pairs of observables that do not commute]"? P0M (talk) 02:17, 3 March 2011 (UTC)

Q and P are thought to be interesting for two reasons. First, because they are fundamental for classical mechanics; second, because they are connected to Fourier transformation. Both issues are irrelevant as far as the physics of (generalized) quantum mechanics is concerned. The point is that applicability of the Hermitian operators Q and P to a generalized bivariate measurement (described by a POVM $\scriptstyle \{R_{mn}\}$) hinges on the compatibility of the observables $\scriptstyle \sum _{n}R_{mn}$ (for different values of m), so as to enable to associate these operators with the spectral representation of a Hermitian operator (analogously for the operators $\scriptstyle \sum _{m}R_{mn}$ for different values of n). In general, however, the operators $\scriptstyle R_{mn}$ are not mutually commutative, and it is a very special case when in a marginal they combine in such a way that the marginal can be interpreted as corresponding to a nonideal measurement of a standard observable (corresponding to a Hermitian operator). Probably this special case is felt to be interesting only because it was presented in our undergraduate courses as some ultimate physical truth. Joint nonideal measurements of incompatible standard observables probably detract from real insight into the meaning of quantum mechanics as much as did the restriction to standard observables (represented by Hermitian operators). It would be interesting, though, to try to see whether real experiments in which particle energy is measured in such a way that we also obtain knowledge of its position can be seen as joint nonideal measurements of energy and position. I never tried such a thing because I feared that measurement inaccuracy would be too large to experimentally reach the quantum limit. I studied other (existing) experiments, many of which turned out to be interpretable as joint nonideal measurements of two (or more) standard observables, however realizing that real progress would probably have to come from more general measurement procedures.WMdeMuynck (talk) 15:16, 3 March 2011 (UTC)

I still need some other pairs.

It should be that any measurement that serves to give "which path" information would, in the naive form of the experiment, simply wipe out the interference pattern, and in some more sophisticated form of the experiment it would "be changed but not completely wiped out." I need some way to make the article not appear to give special relevance to momentum and position. For instance, if memory serves there have been experiments in which polarizers have been placed at the far side of the double slits. One might have vertical polarity and the other horizontal polarity. Under these circumstances, no interference pattern will form. What happens if they are not related at 90° to each other, but at 0°. I think the answer given by experiment would be (has surely been) that no "which path" information is gained and so the interference pattern is not affected negatively. What then would happen if they were related at some angle between 0° and 90°? Surely there is a smooth progression between "no wipe out" and "total wipe out."

If it is not true that any measurement that serves to give "which path" information and which in the naive form of the experiment simply wipes out the interference, and in some more sophisticated form of the experiment will "be changed but not completely wiped out," then something very interesting must be going on. P0M (talk) 20:55, 3 March 2011 (UTC)

I wonder whether the neutron interference experiments we studied in a paper that can be found on the main publications page of my website [[3]] (Publ. 27: Neutron interferometry and the joint measurement of incompatible observables) can be of any use to you. In the paper we analyzed experiments that have been performed in Vienna by the group of Rauch, in which in one path objects are placed allowing a (statistical) determination of the path of a neutron. Note that the path and interference observables are defined by the measurement arrangement. They constitute a simple pair of incompatible observables on a two-dimensional Hilbert space that are measured jointly, be it nonideally. The path and interference observables are determined in eqs (14) and (15) (viz. the projection-valued measures {$\scriptstyle P_{+},P_{-}$} and {$\scriptstyle P_{A},P_{B}$}, respectively).WMdeMuynck (talk) 22:40, 3 March 2011 (UTC)

Possibly, although it doesn't look like it would be easy to make those experiments understandable and relevant to the average well-informed reader. It occurs to me that most if not all of the quantum eraser experiments may potentially involve "partial erasure," and that most or all of the inferometer experiments used to show quantum effects may also be used.P0M (talk) 01:10, 4 March 2011 (UTC)

I agree with what you said here. Seldom if ever reality is kind enough to comply with tthe simple theoretical schemes we are able to think of. In general enlarging the variety of your experiments brings about the necessity to generalize your scheme. I think the enlarged (POVM) scheme of quantum mechanics brings us more closely to classical thinking. But giving up the intriguing conundrums of standard quantum mechanics for a much less exciting scheme is probably not what the average well-informed reader of Wikipedia is looking for.WMdeMuynck (talk) 09:52, 4 March 2011 (UTC)

For instance, the first mirror or fourth could have any degree of reflectivity between 0 and 100% (theoretically, anyway). At any reflectivity other than 50%, the experimenters would know that a certain percent of the photons had to go by way of one path or the other, and they ought not to take part in interference (upper right corner). So, by varying the reflectivity of the first mirror, it ought to be possible to arrange for an interference pattern that is in any desired degree "washed out."

So far it looks like polarization and reflectivity could both be used to demonstrate a selectable degree in interference. Any other mechanisms?P0M (talk) 13:41, 4 March 2011 (UTC)

Experimental arrangements of the type you suggest indeed are useful to demonstrate the possibility of changing the interference pattern by varying certain parameters. Experiments of this type were discussed by us in several papers, for instance, Publ. 31 [[4]] and Publ. 38 [[5]].WMdeMuynck (talk) 20:04, 4 March 2011 (UTC)

O.K. It is clearer what you want, and clearer how the article might reflect this clarification. However, what is not entirely clear to me is why anybody would not see this kind of variability "intuitively," or, to put it another way, what misconceptions people might have that would interfere with seeing something that, to me, seems perfectly obvious. The article currently says, "It is a widespread misunderstanding that, when two slits are open but a detector is added to the experiment to determine which slit a photon has passed through, then the interference pattern no longer forms and the experimental apparatus yields two simple patterns, one from each slit, superposed without interference." This statement is not exactly a misunderstanding, but a correct statement that applies to certain experimental conditions, e.g., the setup with crossed polarizers (90°) vs. the same setup with un-crossed (0°) polarizers.

One justification is given for the assertion that there is a "widespread misunderstanding," a reference to time 3:40 of an on-line video, but there is no analysis of this evidence to show in what way it is in error. It is simply assumed that because there is one video, then that video represents what "widespread opinion" is (and I concede that this assumption could almost certainly be substantiated), and that it reflects a false understanding of what is actually going on in the real world. But the generalized nature of such experiments should behave in a way analogous to that kind of experiment you have described above: "We studied measurement procedures in which a parameter could be changed in a continuous way so as to have the measurement results change continuously from the results of an interference measurement into the results of a which-way measurement." It appears to me that in an experiment in which X percent of singly fired electrons are detected by some measuring device (the video seems to envision shooting photons at electrons and detecting them that way) then X percent of the interference pattern would be "washed out." The video maker must have assumed a total success rate in detecting electrons just beyond the double-slit apparatus.P0M (talk) 08:29, 6 March 2011 (UTC)

Copenhagen Interpretation

The following bit states the obvious 4 times, in different words, that the object is not detected in between emission and detection. In addition if it is "not known" what is going on, then how can additional information be stated like it is "out of sensible interaction with the things of our universe"? Surely a contradiction.

"If by "object A exists" is meant "object A is detected at point x,y,z,t," then this object "exists" only at the point of emission and the point of detection. In between times it is completely out of sensible interaction with the things of our universe, out of sensible interaction with the macro world. What is going on in the apparatus is something that is not known." —Preceding unsigned comment added by 88.5.175.129 (talk) 23:12, 6 February 2011 (UTC)

Photon behavior is not obvious at all, at least not to people who are accustomed to operating in the macro world and keep those expectations in operation when thinking about the quantum world. The human expectation that something will appear at all points along its trajectory seems to be either built in or very quickly learned. Babies have been presented with a contrived scene in which there are two trees or pillars in front of them (in a movie or in some kind of contrived puppet stage perhaps) with a fairly substantial gap between them. Some creature, a rabbit perhaps, walks from one side of the stage up to and behind the first tree. The rabbit then appears from behind the other tree after an appropriate time interval, and continues walking in the original direction. The reactions of infants indicate that they are puzzled by this string of events. The infants used in this experiment were pre-verbal. A person old enough to talk would presumably ask how the rabbit got from behind the first tree to behind the second tree without being seen.

As individuals who have had a fair amount of experience with the world, we do not expect to see birds take off from one tree, disappear, and reappear as they land on another tree. We are accustomed to being able to watch them unless there is something obstructing our vision.

In the case of rifle bullets, we can mark a bullet, put it into a rifle, fire it, and dig it out of a target located some distance away. It does not bother us that we do not see it moving between the two positions because we have learned that things can go too fast to be seen. However, we also know that if we really want to see it in flight we can use high-speed photography.

In the case of photons however, we can do something analogous to loading a rifle and firing it. We can activate a laser device that is designed to emit one photon at a time. We can do something that is analogous to digging the bullet out of a target because the photon will activate a CCD photon detector or make a spot on photographic emulsion. But in between we can do nothing to spot the photon the way we can spot a flying bullet.

The only time that we have knowledge of the photon are at the two times noted -- when it is produced by a laser we know where and when it came from, and when it is detected we know when and where it showed up. In between these two times, anything we say about the path taken by the photon is a matter of conjecture and logical reasoning about what it "must have done." Richard Feynman says that that photon has traveled by "all possible paths" between origin and destination. In the double-slit experiment, people argue about which of the two slits the photon has "really" traveled through.

"Out of sensible interaction" and "being unknown" are, to me, not contradictory statements, but virtually tautological statements.

If someone says, "I know where the photon is at the midpoint of its travel," what does that person mean? In other words, what is the operational definition for "knowing where something is"? I do not know any scientifically acceptable answer to a question such as, "Is your passport still in your breast pocket?" except to check your breast pocket. You may have put it there, and you may remember correctly that you put it there. Perhaps it has fallen out, or perhaps a pickpocket got it. The only way to be sure is to look. P0M (talk) 05:34, 7 February 2011 (UTC)

I found this entry misleading and biased - without a fair representation of the significant controversy over the role of the observer in the collapse of photon or electron particles. jamenta —Preceding unsigned comment added by 98.155.237.197 (talk) 03:54, 8 February 2011 (UTC)

I presume "this entry" does not refer to my remarks immediately above those words, but to the article itself. Can you point us at a coherent discussion offered by credible/authoritative specialists in the field who believe in the necessity for there to be a conscious observer involved to make an indeterminate state "collapse" into a determinate state -- or a discussion by some recognized physicist who expresses whatever idea it is that you find missing in this Wikipedia article? P0M (talk) 18:54, 8 February 2011 (UTC)

Without necessarily referring to a conscious observer, John Wheeler did write, "We ask the yes-or-no question, 'Did the counter register a click during the specified second?' If yes, we often say, 'A photon did it.' We know perfectly well that the photon existed neither before nor after the detection." This implies that a photon's very existence is inextricably tied to the detection event. I don't know if that is relevant to the above editor's objection, but Wheeler was a heavyweight who subscribed to some form of observer-participancy. -Jordgette (talk) 00:42, 9 February 2011 (UTC)

I am saying exactly what Wheeler says. Something was presumably going on before the counter registered a click during the specified second, but such "goings on" are not observed. If the counter had been installed in a lightproof vault, and a single-shot laser was put in there aimed at the counter, we would be upset if the counter occasionally clicked without our energizing the laser immediately beforehand. We would presumably look for leaks. If we didn't find leaks we would probably put a parasite voltage meter into the laser aparatus in parallel with the circuit that is immediately responsible for setting off an emission just so that we could be sure we weren't getting some kind of accidental firing of the laser. If the detector started receiving in morse code we would start looking for pranksters or ET. My point is that we are extremely suspicious of anything that appears to be an uncaused event.

Now if the operational definition of "to exist" means "found proof of whatever it is at x, y, z, t," then the photon does not "exist" until it hits the detection screen attached to the counter. We could stick another detector somewhere along the original path between the laser and the original detector/counter, and by that means we could "make the photon show up" or "make the photon come into existence" at an earlier time and a different place.

Look at it this way: If we tested the original sealed vault device thoroughly, and got used to seeing a "ping" on the wire leading to the outside world from the counter whenever we fed in a signal to the laser to fire, we would be happy with our experimental device. I am not sure whether anybody would seriously consider the possibility that the external device would not ping if there was nobody there to watch it, but now consider another wrinkle:

We modify the original device simply by putting a detection screen midway between the original components, but we don't bother with a wire to the outside, and in fact we don't even bother with a counter. If photons only come into existence when somebody observes them, then the unobserved detection screen should have no influence on what shows up or comes into existence on the only detection screen where it should be possible to observe it.

Bohr was clear about what "to observe" meant in these circumstances, and he explicitly denied that it had anything to do with human consciousness.

It's a good thing that photons do not require a human observer to manifest themselves. Nothing would have happened in the universe until humans evolved, and humans could not have evolved until things started happening. And humans would have to be everywhere and everywhen to make the poor universe come of age.

We have to have photons "showing up" before we can observe them, but that truth does not imply that we have to observe photons in order that they may show up.P0M (talk) 03:34, 9 February 2011 (UTC)

I find P0M's description to be a well-written, straight-forward way of helping readers relate to the weirdness of the quantum world. What do others think of incorporating something like this into either this article or Introduction to quantum mechanics? —UncleDouggie (talk) 01:34, 9 February 2011 (UTC)

Something like the current explanation of the events is required (ie, from the Copenhagen interpretation). It makes clear the fundamental implications of this experiment, which are less obvious from reading the introductory section of "Quantum version of experiment". I think however that the language could possibly be unrelaxed, as the phrases highlighted as requiring citations appear biased to an extent by the way in which they are written; eg "it is perhaps not so astounding". It may also be less obvious to a reader and that "universe" here refers to that which can (at least in theory) be observed or "known", ie the physical universe. I am also not sure if this is an appropriate section to start entering into debate regarding the definition of an observer; the presence of an observer is an assumption built into the interpretation Richardbrucebaxter (talk) 11:27, 2 April 2011 (UTC)

Request for an answer be put into the wiki page

I wanted to know: if two detectors were used to detect which slit the photons went though but the result of this detection is not viewed/recorded, then what would be pattern? interference or not?

There are a lot of discussions in forum, but without expert option, we are not sure. I believe the answer is no interference pattern.

The current wiki page has the following text: "Even less in line with the expectations of human scale interactions with nature, if the information about which slit a given particle came through is "erased" before a photon has time to interact with the detector screen, interference will be restored"

I think others are mis-understanding this. They think no viewing or no recording implies the which-path-information has been erased and therefore an interference will return.

Please update the wiki with a correct answer; if you think this question is worth answering.

WhoperJ12 (talk) 19:33, 12 April 2011 (UTC)

The article could be clarified in this regard. You're right, interference in this case diminishes regardless of whether the which-path info is consciously viewed or recorded. The determining factor seems to be whether the which-path info is knowable 'in principle' — whether or not this information merely exists in the world. For the interference to be restored, the which-path info needs to be eliminated entirely ('erased') so that the particle's path could never be determined (by a person, detector, etc.). It would be great if someone had a good source for this; perhaps it could be added to the article. This question does seem to keep coming up again and again, and it tends to send some people down the wrong path. -Jordgette (talk) 22:48, 12 April 2011 (UTC)

Basically, here's the deal: A wavefunction goes through both slits. The two parts if it interfere with each other. When the two of them reach the detection screen (where the vast majority of them will show up), the photon will show up somewhere on the screen. Then another wavefunction goes through both slits, etc., etc., and another photon will show up somewhere on the screen. I don't know who is keeping score, but the "hits" on the screen always follow the probabilities that the wavefunctions determine. (You want them all to have the same frequency so that their probabilities of where to show up will follow the same rule. White light doesn't work well because the "hits" for each frequency fall by different rules.) So you get a nice interference pattern after enough hits have accumulated. (The video of the electrons hitting a screen one by one and gradually forming an interference pattern is a great way to see how the interference pattern is formed.)

Now what would happen if we put in polarizers just beyond each of the two slits, and we rotate them 90° to each other? One waveform is spread out vibrating left to right across the screen, and the other waveform is spread out vibrating top to bottom across the screen, so they do not interact at all. In that case all we get are a couple of diffraction patterns. When we say that we have "marked" the photons as to which path they took, we have not reached out and put a tag on each photon. What we have actually done is to rotate half the waveforms. We could set up this experiment using a laser that only fires off one photon at a time, and we could put a second polarizer farther down the path from the laser to the screen. It would match the polarity of one or the other polarizers nearer to the double slits. Then if a photon showed up on the detection screen, we could argue that having gone through the second polarizer, and the second polarizer having the same orientation as the polarizer on the left slit, the photon must have gone through the left slit. So the way we mark a photon is in this case a little like the way we could mark sheets of paper for special attention just by rotating them 45° to the papers that don't need special attention.

Having messed up the interference patterns by rotating each photon's two part (from two slits) wavefunction so they don't coincide with each other on the detection screen and so can't interfere, how, short of taking out the polarizers, can we get things so that each waveform finds both of its parts vibrating in the same direction? The answer is to use another kind of polarizer that will give half of the wavefunction pairs a diagonal polarization and the other half of the wavefunction pairs a counter-diagonal polarization.

The above is all pretty abstract. Let me see whether I can use some simple symbols to serve in the place of an SVG image.

In the beginning, the wavefunctions that came through the double slits all were vibrating in the same direction. So we can imagine their schematic diagram being a sort of squashed "=" sign. The wavefunction part coming from the left slit would be the top line, and the wavefunction part coming from the right slit would be the bottom line. But actually there would be no vertical space between them.

With the polarizers in place, the mode of vibrations of the two-part wavefunctions are 90° rotated from each other, so a schematic would look sort of like a "+" sign. If somebody puts another 90° polarizer into the apparatus, we get either "|" or else "—". We have effectively prevented one or the other part of each wavefunction from getting through, so there is no way in all of this (with two polarizers or three polarizers) that we can have interference.

However, we can use another kind of polarizer as the third one. It is one that takes "|" vibrations and changes half to "/" and the other half to "\". It also takes "—" vibrations and changes half to "/" and the other half to "\". It does so in a coordinated way, so that half of the wavefunction pairs that formed the original "=" arrangement are, as it were in a "//" arrangement and can interfere, and the other half of the original wavefunction pairs are in a "\\" arrangement and can interfere. If I remember correctly, the "X" arrangement doesn't work perfectly because the "\" part and the "/" part are displaced from each other.

The result is a little messy, but by successive operations we have managed to get things back to a situation in which they can interfere with each other.

The term "erased" is unfortunate. Nothing is really written, and nothing is really taken off with a scraper of some kind. A physical change of some kind is done to the wavefunctions so that they do not get superimposed, and then later something else is done so that they will again be superimposed.

The experiment with polarizers was the subject of a Scientific American article, which is cited in one or more of the articles on quantum "erasure." I've tried to give you the gist of it here.

As far as I know, "erasure" is always like this. For instance, you might be able to delay the arrival of whatever goes through the right slit, so that it would not be around to interact with whatever goes through the left slit. Time of arrival on the detection screen would tell you which slit the photon went through. But that "marking" of half the wavefunction pairs flowing to the screen could be compensated for by delaying the stuff going through the other slit. By the way, nothing anybody does can "erase" a result or an action after an individual photon has shown up on the detection screen. By "marking" and "erasing" is meant something like this: "Originally I had a nice interference pattern. I do X to differentiate the left slit stuff from the right slit stuff, and the interference pattern disappears. But if I do Y then the interference pattern comes back again." Of course the interference pattern is like the river that you never step into twice.

I'm working from memory of some of the erasure articles. I'm pretty sure that one of them has diagrams that show the physical difference between an interference pattern produced by the "/" vibrations and an interference pattern produced by the "\" pattern. I'm not optimistic, however, about finding an article that lays things out for somebody who doesn't already know the details. Most writers use the shorthand catch phrases like "erasing the which-path information" because they know that their professional colleagues will understand the physical operations they are talking about.

Human beings are tiny creatures in an immense universe that has been around far longer than we have. It does not matter what we see, what we are aware of, what we are conscious of. We are not that important. As far as I know, nobody now believes that if we set up a double-slit experiment in a bank vault with no observers allowed, and recorded the results on the detection screen in some permanent form (perhaps we had an electronic device to copy the detection screen to an engraving on some metal that would not corrode), and we left it there even for thousands of years, the Universe would not wait until somebody cracked the time vault to play catch-up and record the pits into the metal plate so that the arriving humans could have a nice interference pattern to see.P0M (talk) 01:38, 13 April 2011 (UTC)

Contradictory affirmations?

The article lead reads "Any modification of the apparatus that can determine which slit a photon passes through destroys the interference pattern, illustrating the complementarity principle: that light (and electrons, etc.) can behave as either particles or waves, but not both at the same time." and then goes on to say "However, an experiment performed in 1987 produced results that demonstrated that information could be obtained regarding which path a particle had taken, without destroying the possibility of interference." Don't these two affirmations contradict each other? I modified the first sentence to read that it was originally believed to be so, but I got reverted. --uKER (talk) 16:08, 14 April 2011 (UTC)

Well, I just noticed that this discussion deals with that very same thing. --uKER (talk) 16:11, 14 April 2011 (UTC)

Recent changes by DJr

An editor wanted to reword the following passage:

Any modification of the apparatus that can determine which slit a photon passes through destroys the interference pattern,[5] illustrating the complementarity principle: that light (and electrons, etc.) can behave as either particles or waves, but not both at the same time.[6][7][8] However, an experiment performed in 1987[9] produced results that demonstrated that information could be obtained regarding which path a particle had taken, without destroying the possibility of interference. This showed the effect of measurements that disturbed the particles in transit to a lesser degree and thereby influenced the interference pattern only to a comparable extent.

The revision didn't work, but the current passage is not clear. The discussions I have seen have been short on context because they are written for professionals who already know the background. We need to improve this paragraph to show, in more detail, how these experiments work and what they actually measure and do not measure.P0M (talk) 18:22, 14 April 2011 (UTC)

Having watched this discussion for a long time, this is what I propose:

Any modification of the apparatus that can determine which slit a photon passes through weakens the interference pattern,[5] illustrating the complementarity principle: that light (and electrons, etc.) can behave as either particles or waves, but not both at the same time.[6][7][8] An experiment performed in 1987[9] produced results that demonstrated that information could be obtained regarding which path a particle had taken, without destroying the interference altogether. This showed the effect of measurements that disturbed the particles in transit to a lesser degree and thereby influenced the interference pattern only to a comparable extent.

-Jordgette (talk) 02:10, 15 April 2011 (UTC) MartyP 16 May 2011 12:24GMT Hey boffins, interesting stuff. Sounds like a photon is a soap bubble (travelling at light speed) that bursts and creates a wave if it touches anything. —Preceding unsigned comment added by 194.125.96.83 (talk) 11:26, 16 May 2011 (UTC)

hair deletion

This paragraph was just deleted with the comment that it is irrelevant to the double-slit experiment:

- Interestingly, this same equation works when there are no slits at all in i.e. a "zero slit experiment" akin to the one performed by Thomas Young or Isaac Newton. If one shines a laser pointer at a human hair in a dark room, the interference fringes on the wall will be the distance "x" apart. The hair is ~100 micrometres thick, plus or minus about 80 micrometres depending on the hair, and is the variable "d". A cheap laser pointer emits light at wavelengths ~660 nm. One can adjust L by walking toward or away from the wall.

It was inaccurate to describe this experiment as a "zero slit" experiment. It is essentially the same experiment done by Thomas Young. He used a "card" if my memory is correct. The "inner edges" of the two slits are there, but the outer edges have disappeared. One could sneak up on the conditions of this primitive form of the double slit experiment by making slits with the same central solid zone but with increasing slit width. P0M (talk) 16:04, 17 May 2011 (UTC)

It is not a zero-slit experiment, it is a single slit experiment. The form of the diffraction is the same as would be obtained with a slit of the same width - see Babinet's principle. There are no double slit intereference fringes, only diffraction minima and maxima - though as Feynman commented, there is no well defined difference between interference and diffraction. But the important point is that double slit interference has one set of fringes whose spacing is related to the separation of the slits and whose intensity vasries as the square of the sine, upon which are auperimposed the diffraction pattern of the individual slits which is a 'sinc' squared function.

In the far field case- seeFraunhofer diffraction, for a single slit, or a hair of the same width, we get:

$I(x)\propto \mathrm {sinc} ^{2}\left[{\frac {2\pi a\sin \theta }{\lambda }}\right]$

apart from an additional forward intensity from the undiffracted beam in the case of the hair.

while for the double slit case, we get:

$I(\theta )\propto \left[{\frac {\sin {(ka\sin \theta )}}{ka\sin \theta }}\right]^{2}\cos ^{2}{[k\sin \theta (d+a)]}$

the width of the aperture(s) is 2a, and the separation of the slits is d. Epzcaw (talk) 08:33, 20 May 2011 (UTC)

On reflection, however, you could also consider the output waveform as arising from the interference between the two beams formed by the hair spliting the laser beams in two, and the parts which are 'diffracted' by the edges away from their original direction giving rise the the intensity variations which could hten be labelled 'interference fringe' rather than a 'diffraction pattern'. Maybe a matter of semantics. You could then also label the pattern you get when you send light through a single slit as being caused by the interfernce of the light 'diffracted' by the edges, and therefore label this as an interference pattern as well.

In reality, light is not 'diffracted from the edges' - all points of the wavefront are spreading out all the time, but when you remove bits of the wavefront, the sum of the spread-out wavefronts changes.

Might be more appropriate to have this in a new section called 'Diffraction and Interference'. I'm not sure it is helpful in this section which is primararily directed at discussing wave-particle duality,

But up to you.... Epzcaw (talk) 08:43, 21 May 2011 (UTC)

recent addition by an unregistered editor

Somebody recently added:

When a detection device is placed in front of the slits to observe incoming particles one by one, the observation collapses the wave function of the incoming electrons, and they pass through behaving like particles, creating two bands instead of an interference pattern. When no such observation is made, the wave function is not collapsed going into the slits, and the wave function potentials interfere, producing interference patterns.

What the expression "in front of" means is dependent upon where the reader imagines himself/herself to be standing. P0M (talk) 02:53, 18 May 2011 (UTC)

Deleting unclear/unsubstantiated paragraphs

I deleted the following paragraph:

"Restriction to the two experiments in which either both slits are open or one slit is closed has given rise to the idea of wave-particle complementarity according to which a microscopic object (photon, electron, etc.) would manifest itself as a particle in the which-way experiment but as a wave in the interference experiment. This idea has been felt to be counterintuitive by those not being content with an instrumentalist interpretation of quantum mechanics in which that theory is accepted as just describing phenomena without providing explanations."

First of all, the paragraph is poorly written and doesn't define terms like "which-way experiment". Also, if you look at the page for the instrumentalist interpretation, nothing is cited, and the article is incoherent. I don't think fringe theories belong up in the overview section. Eratosthenes (talk) 09:10, 23 May 2011 (UTC)

Newton's double slit experiment

The first section of this article says

"The setup used by Thomas Young, and by Newton, differs from the modern version; they passed a beam of light over a thin object such as a slit of card (in Young's case) or a hair (in Newton's case)."

I think it is very misleading to mention Newton's experiments with light falling on hairs in the context of the double slit experiment. Newton was looking for diffraction effects which had been described by Grimaldi, and were demontrating what is now called diffraction in wave theory. Newton, however, did not accept wave theory, but proposed that light was made up of particles which are emitted by light.

Young's double slit experiment was extremely important in the development of the understanding of light and was one of the keystones in establishing wave theory which is still a valid model. I think it would be better to remove the reference to Newton's work here. Epzcaw (talk) 09:58, 3 June 2011 (UTC)

Unregistered user changing article evaluation

I have reverted a change by an unregistered user once. I don't know who is supposed to be evaluating articles, but the changes should not be made on the personal whim of somebody who visits the page once or twice. The user changed the evaluation a second time. I've left it for others to deal with.P0M (talk) 17:17, 26 June 2011 (UTC)

I agree with you that those assessments should not be made in a whim but I also agree that this page should be assessed as "top importance". Dauto (talk) 18:25, 26 June 2011 (UTC)

It didn't used to be "start class," no? P0M (talk) 21:20, 26 June 2011 (UTC)

It was class B. I fixed it. -Jordgette (talk) 22:55, 26 June 2011 (UTC)

If this article is B class, it suggests Wikipedia has very poor standards. It is unstructured, incoherent, repetitive. It has two sections which lacks citations and one is criticised for its personal reflection style. I think it would be if very little help to someone trying to understand what the significance of the experiment is in physicsTreadai (talk) 10:14, 27 June 2011 (UTC)

I have compiled some detailed criticisms of the article here:

The introduction reads as if it was written by several different people (which presumably it was) as it repeats itself. For example:

Para 2: “In quantum mechanics the double-slit experiment demonstrates the inseparability of the wave and particle natures of light and other quantum particles (wave–particle duality).

Para 5: “The double slit experiment can also be performed (using a different apparatus) with particles of matter such as electrons with the same results, demonstrating that light and matter have both particle-like and wave-like properties (wave–particle duality).”

Or:

Para 2: The wave nature of light causes the light waves passing through the two slits to interfere, creating a pattern of bright and dark bands on the screen. (However, at the screen the light is always found to be absorbed as though it were composed of discrete particles, photons

Para 3:With light, although in many circumstances it behaves as particles (photons), it has been known for over two centuries that the pattern with two slits is not the sum of the separate patterns—this established the wave nature of light. The actual distribution of brightness can be explained by the alternately additive and subtractive interference of waves

The introduction and the Overview says contradictory things:

Introduction: “Any modification of the apparatus that can determine which slit a photon passes through weakens the interference pattern,[3] illustrating the complementarity principle: that light (and electrons, etc.) can behave as either particles or waves, but not both at the same time. An experiment performed in 1987[7] produced results that demonstrated that information could be obtained regarding which path a particle had taken, without destroying the interference altogether”

Overview. “The addition of a detector to the apparatus does erase the interference, so that the probability density curves add separately without any extra terms for interference.”

Underpinning of the experiment - what does this mean????

What is the relevance of “triple slit experiment”, “Shape of the fringes” ?

The section "Photons travelling through two slits” is a repetition of what is in the overview, but phrased differently, which is very confusing to a new reader.

The “Importance to Physics” is mainly a repetition of the “Overview” section with no expansion.

Overview: “The experiment was first reported by Thomas Young, (strictly speaking, it was a double hole rather than a double slit experiment) and it played a vital part in the acceptance of wave theory of light in the early 1800s, vanquishing the corpuscular theory proposed by Newton which had been the accepted model of light propagation in the 17th and 18th centuries.”

Importance to Physics: “Although the double-slit experiment is now often referred to in the context of quantum mechanics, it is generally thought to have been first performed by the English scientist Thomas Young in the year 1801 in an attempt to resolve the question of whether light was composed of particles (Newton's "corpuscular" theory), or rather consisted of waves traveling through some ether, just as sound waves travel in air. The interference patterns observed in the experiment seemed to discredit the corpuscular theory, and the wave theory of light remained well accepted until the early 20th century, when evidence began to accumulate that seemed instead to confirm the particle theory of light”

Overview: “Richard Feynman was fond of saying that all of quantum mechanics can be gleaned from carefully thinking through the implications of this single experiment”

Importance to Physics: The double-slit experiment, and its variations, then became a classic thought experiment for its clarity in expressing the central puzzles of quantum mechanics.” Quantum version of the experiment - this meaningless. The experiment is the experiment. There is a classical wave optics interpretation, and a QED interpretation. Again this section repeats much of what went before.

I could go on and on, but if this was work from a student, I would have stopped well before this point and asked for a re-write.

I don't propose to alter the quality rating, as no doubt it would be reverted. But those people who think it merits quality class B should take a close look, and perhaps improve it. At least, the already added suggestions about citations, and style should be attended to. Treadai (talk) 16:27, 27 June 2011 (UTC)

B-class refers to articles that are more or less complete, with no major gaps in information, but still need organizational and style work. To me it sounds like that describes this article pretty well. It certainly needs improvements, but C-class is described in part as "still missing important content or contains a lot of irrelevant material" and "Considerable editing is needed to close gaps in content and address cleanup issues." I don't think it's so disorganized as to merit a C. It certainly isn't Start class. -Jordgette (talk) 18:21, 27 June 2011 (UTC)

Clearly a difference of opinion then!!

If a friend wanted to know what the double slit experiment was, and why it is important in physics, I certainly would not suggest they read the Wikipedia article, as I think it would put them off the topic for life.

It is interesting to look at the history of the quality rating of this article. Up to November 2006, it was not rated.

On 18 Nov 2006, LINAS rated it as Start class without comment.

On 18 Jun 2008, HEADBOMB changed the rating to A class with the comment - "Looks A class to me (Glanced at it for at least 15 seconds. At least B)".

On 23 Feb 2009, DRKIERNAN changed it to B class without comment.

On 26 Jun 2011, an unregistered user changed it to Start class without comment.

What this says to em is that the quality rating system is meaningless, and so I won't waste any more time on it.::I do think that because of the way in which Wikipedia works, articles need to be re-written from scartch every so often, obviously including the good bits already there. I think this needs to be done by a few people who are willing to be both crtiical and criticised. So up to you folks.Treadai (talk) 09:47, 28 June 2011 (UTC)

I agree that the article needs a top-to-bottom reworking. Thank you for pointing out many of its rough spots. -Jordgette (talk) 18:23, 28 June 2011 (UTC)

True nature of ' matter', interference of waves produced by 'free-floating-antennas'

For more than twenty years I tried to communicate the true nature of 'matter' through the system of 'peer-reviewed-journals', but the prestigious journals are not courageous-enough to publish. So I am expressing it here to the open-minded readers of dis page: 'particle' is not a 'substance', 'particle' is a 'process' i.e. 'a phenomenon' of constructive-superimposition of very wide band of pre-quantum-mechanical-real-waves. 'Matter' is a process of fluctuations taken place in a 'continuum'. The continuum nature of the most-fundamental-reality allows formations of spherical 'wave-packets' of micro-microscopic-dimensions. There are integer, whole number of such 'wave-packets'. The constituent-waves of these packets sprade in all the directions. An interesting difference between the interference of conventional electromagnetic waves and the waves of 'matter-particles' is: that in the case of electromagnetic waves generated by radio-stations, the waves add constructively or distructively depending upon their relative phase, and the antennas remain firmly fixed; whereas in the case of 'matter-waves' depending upon the constructive or distructive superimposition of the waves, the antennas change deir positions! Because the 'particles of matter' are so light-weight, that they are like 'free-floating-antennas'. Interefrnce of 'the most fundamental waves', generated in the continuum, causes changes in the positions of emergence of the 'spherical wave-packets' called 'particles of matter'. We require not just 'strings' or 'loops' but rather three or four-dimentional-continuum to describe the 'particles of matter'. Thus, the particles of 'matter' are 'particles' as far as their micro-microscopic size, and their whole, intiger number is concered; and they are 'waves' as far as their true nature of 'fluctuations of the most fundamental continuum' is concerned. Hasmukh K. Tank, 22/693 Krishna Dham-2, Vejalpur, Ahmedabad-380051 India 122.169.65.146 (talk) 13:52, 28 June 2011 (UTC)

I have overhauled the article

So I've gone through and done a massive restructuring of this article, currently in my userspace. I tried to resolve most of the contradictions and repetitions that have been pointed out, and I deleted some of the images and text that seemed more appropriate for Young's article. I got burned out, though, and didn't line-edit anything beginning with the maths section, so there may still be some repetition. Overall it got quite a bit shorter. Please check it out and give me feedback on anything that still might be less than clear, confusing, or repetitious. Thank you. -Jordgette (talk) 00:57, 29 June 2011 (UTC)

I just read it, and it looks fine to me. There may be a couple places where a little longer explication may make for less puzzlement on the part of the reader, but maybe I am just slow. I did fix one typo, but left your draft alone otherwise.Thanks for the good work. P0M (talk) 03:00, 29 June 2011 (UTC)

I have looked at your revised article, and it looks excellent to me - vast improvement(I need to spend a bit more time to be able to nit-pick on details!!!).

I have proposed before (see above) to remove the classical wave explanation to a separate article -I'm not sure now if this is appropriate or not. If I get round to doing it, I will first check with you both (as you two seem to have a major input into this article).

Another suggestion - have you seen a recent article in the European Journal of Physics by Giorgio Matteucci -'On the representation of wave phenomena of electrons with the Young-Feynman experiment' (32, no 3 p733). He makes the point that single slit/hole diffraction is just as good a demonstration of the wave-nature of electrons as the double slit experiment and provides some rather nice pictures of electron diffraction by holes. Billiard balls passing through a single hole would not show diffraction rings, so when you first start thinking about quantum-mechanical particle, these results are just as baffling as the double slit experiment. Or maybe this should be another article?

Will go through the article in detail later, and for quality control checkEpzcaw (talk) 08:58, 29 June 2011 (UTC)

I think single-slit diffraction should go in its own article. P0M (talk) 13:55, 29 June 2011 (UTC)

I have copied your article into a sub-user page and made comments (mainly minor) in bold at http://en.wikipedia.org/wiki/User:Epzcaw/Double_slit#Overview

Epzcaw (talk) 17:58, 29 June 2011 (UTC)

Definitely B now. Maybe B+ Treadai (talk) 14:46, 30 June 2011 (UTC)

So where are we now?

I've lost track of developments on this article. The last time I noticed any serious suggestions, we had two alternatives in two user-spaces. We seemed to be making progress, but when I look for changes in the article history I don't see indications of big changes. Am I missing something?P0M (talk) 21:31, 6 July 2011 (UTC)

I agree. Jordgette talk re-wrote the article in his/her user space to what I (and I think you) thought was much improved, but hasn't done anything further since the end of June. Maybe contact him/her? Epzcaw (talk) 08:34, 7 July 2011 (UTC)

Sorry folks, I haven't been able to put everything together. I'll probably do it next week, and if there are any changes to the article in the meantime, I'll incorporate them. -Jordgette (talk) 17:45, 7 July 2011 (UTC)

 Done I put all of the changes together and made a few more. For example, I found it confusing that we start off saying the experiment demonstrates wave/particle, but then we usually refer to the quantum as a particle (e.g., which slit did the particle go through). So I clarified that at the top of the individual particles variation, and expanded the intro to the particle-detector variation.

As a couple of people have noted, the intro to the maths section still needs work. It can be simplified further. What I'd like to have is a simple diagram showing flat wave fronts emerging from a single slit, and then from a double slit. Seems like that should be around but I couldn't find it on WP. Then perhaps we could get in Epzcaw's point above about the single slit. Anyway, take a look. -Jordgette (talk) 04:41, 10 July 2011 (UTC)

Is it accurate to have flat wave fronts emerging? The wave fronts that emerge from a laser are not flat, but a lens put in front of the laser can remove the curvature. When a wave emerges from a slit, however, it will not be flat -- else there would be no diffraction. P0M (talk) 04:54, 10 July 2011 (UTC)

My mistake, I meant flat wave fronts approaching and spherical wave fronts emerging. -Jordgette (talk) 19:43, 10 July 2011 (UTC)

Looks very good now. A few minor points.

It might be better to make the first particle link to this one which is a more classical one, or or here since the perceived problem is the discrepancy between waves and classical particles.

In the section "With particle detector", the phrase "weakens the interference pattern" needs to be defined. Does it mean that the fringe visibility is reduced?

"Three slits" - I have no idea what any of this means - what "is reduced to a number which is almost zero". Classically, a three slit interference is no more special than a quadruple slit one or a diffraction grating. Unfortunately, the only on-line link is in German. Maybe needs expansion as separate article, or removal?

"Copenhagen interpretation section" - requirement for observation of interference - shouldn't this say "that there should be two separate paths via which the particle can travel" rather than "that there be a screen with two slits" since many of the experiments covered here are not really two slit experiments, but two path experiments?

Classical explanation need simplification. I planned to do something on this - will maybe start now.

Single slit stuff - I agree with POM now - I think this would be better in a separate article. The double slit experiment has a special place in the history of thinking about wave-particle duality, even though what is happening is no more odd than getting diffraction maxima and minima with a single slit (is the particle interfering itself to to do this), but I think its special place needs to be respected.

Would it be worth pointing out the similarity between the path-integral formulation and the Huygens-Fresnel principle?Epzcaw (talk) 11:56, 10 July 2011 (UTC)

I have written a section called "Classical wave optics formulation" in my user page. If you are both ok with this, I will substitute it for the current equivalent page.Epzcaw (talk) 16:45, 10 July 2011 (UTC)

It's much better. I made most of the above changes, although I balked at the triple-slit section -- no one likes it much, but I'm uncomfortable deleting it, partly because I don't quite get it either. I'll leave it to someone else. -Jordgette (talk) 20:28, 10 July 2011 (UTC)