Talk:Wigner's friend

Superposition
Can someone link this word in the article to what it means? Eg is it Quantum_superposition. I am trying to understand this page but am stuck on superposition. Crackerlacken

—Preceding unsigned comment added by Crackerlacken (talk • contribs) 03:48, 9 January 2008 (UTC)

I have some ideas:

1) Instead of a friend, have the machine produce either a 2 000-gram or a 2 001-gram bag of sand. You cannot tell which the bag is merely by picking it up... but will picking up the bag collapse the wave function? Will weighing the bag on an accurate scale collapse the wave function?

2) Have the machine produce a bag containing either 50 003 or 50 008 ball bearings (of non-uniform weight). Will (a human) counting the ball bearings collapse the wave function? If it is known that the person who counts the ball bearings tends to make errors in counting, will his attempt at counting still collapse the wave function? Will merely looking at each of the ball bearings in turn, without (consciously) counting them, collapse the wave function? Suppose we have a machine count the ball bearings. Let us say that it is a mechanical device, with a mechanical counter. If you look at all of the digits of the counter except the last digit, will this collapse the wave function? If you look at the right-hand half of the right hand digit (since "3" and "8" have similar-looking right-hand halves), will this collapse the wave function?

Number 2 is not at all what I call farfetched. I have, at times, tried counting things several times, and got different answers. I began to honestly wonder if the number of (macroscopic) items in a bag had to be "fixed", or if it could be "fuzzy" in a quantum sort of way.

"Alive cat" as direct quote?
There is the line "alive cat/happy friend,". Was the term "alive cat" actually used in any source material? Typically English treats "alive" as a predicate adjective which may not precede a noun; we would rather say "the living cat" or "the cat is alive", but not "the alive cat". samwaltz 19:18, 18 January 2007 (UTC)

Amend second par for clarity
The following par was the 2nd one of this article. I have replaced it with the one below it, for reasons given at the end.

"It posits a friend of Wigner who performs the Schrödinger's cat experiment while Wigner is out of the room. Only when Wigner comes into the room does he himself know the result of the experiment: until this point, was the state of the system a superposition of 'dead cat/sad friend' and 'alive cat/happy friend,' or was it determined at some previous point?" "It posits a friend of Wigner who performs the Schrödinger's cat experiment after Wigner leaves the laboratory. Only when he returns does Wigner learn the result of the experiment from his friend, that is, whether the cat is alive or dead. The question is raised: was the state of the system a superposition of 'dead cat/sad friend' and 'live cat/happy friend,' that was only determined when Wigner learnt the result of the experiment, or was it determined at some previous point?"

1.“he himself” is confusing. It is not immediately apparent whether the pronoun refers to Wigner or his friend. I have made “he” unambiguously refer to Wigner.

2. I have made it manifest that Wigner learns the result of the experiment from his friend, not from direct observation.

3.I have remedied stylistic awkwardness of last part of sentence.

4.I have substituted “live cat” for “alive cat”. Myles325a 02:29, 17 August 2007 (UTC)

What about the consciousness of Wigner's friend?
An element of this thought experiment that I found puzzling is that it is Wigner's friend who first learns of the result of the experiment. If thought experiment is designed to illustrate that consciousness has a privileged function in collapsing the wave function, then is it not Wigner who collapses it first? In that case, Wigner has come too late to collapse it himself.

Otoh, suppose Wigner's friend had no idea of what the experiment involves. He only has to push a hutton, and tear off a print-out produced automatically by a printer attached to the box, upon the completion of the experiment. This printout contains a coded message detailing the result of the experiment. In such a case, we might posit that if Wigner's friend can read and understand the message, the act of reading collapses the wave function and the cat can be said to definitely dead or alive. But if he cannot, then it is Wigner himself who collapses the wave function when he returns and reads the printout himself. Myles325a 02:59, 17 August 2007 (UTC)

Recent Rewrite
The recent re-write is not good. It is completely idiosyncratic and reads like a personal essay. 1Z (talk) 22:17, 10 January 2008 (UTC)

What the hell is a bad world count?
I'm not asking for a lecture on MW. I've been studying it for years, and I'v eenver heard the term. 1Z (talk) 22:31, 10 January 2008 (UTC)


 * A "bad world count" is a nonstandard term. But for some reason I know what the author was getting at. Imagine doing a spin detection on a particle in the following unnormalized state:
 * $$( |up\rangle + 2 |down\rangle)$$
 * The end result is |Saw UP> + 2 |saw DOWN>. Now if you start "world-counting", meaning, counting the number of different observers that this superposition describes, you disect the wavefunction into two worlds. So if you naively think about it as a splitting of consciousness or whatnot, then you would expect a 50% probability. But quantum mechanics says you |saw DOWN> 80% of the time. So the bad world count gave a wrong probability.


 * So maybe things get better if you do the experiment again and again. But if the initial state is a product of (up +2down)(up + 2down) etc and you do the experiment N times, the counting of worlds gives you the same distribution of observers, a 50/50 split. It's wrong.


 * Everett realized that this is not the right way to count worlds, and he introduced a measure on the worlds--- the idea is that worlds with small wavefunction values count less, and worlds with very small wavefunction values count hardly at all. If you add this assumption, then you can show that the worlds where the splitting is different from 80/20 have a negligible weight. But this definition requires you to weight the experiences of observers so that the probability comes out right. There is only one reasonable way to do this, as shown by Everett and Wheeler, because the continuity assumption that small amplitude leads to small likelihood determines the whole psi squared probability. So it's ok. Its philosophically troubling to some people at first, because they might believe that anytime the wavefunction leaks into a tiny corner of phase space a "real" observer pops into being.


 * I've heard thirdhand that there were arguments for the uniqueness of psi-squared probability by Bohm and others in the previous generation, along similar lines except without duplicated observers. It isn't difficult to translate the argument into Copenhagen, except there you would use a detector measuring multiple times instead of an observer looking multiple times, but the result is the same. At the end of N measurements on identically prepared systems, you measure the detector, and then you find that whatever your psi-to-probability measure (so long as zero psi gives zero probability and it is continuous), for large N it becomes certain that the detector detected 80/20, confirming that the psi squared measure is the correct one.


 * Getting the world count right is the philosophical headache in Everett style interpretations. It's essentially the same difficulty as the interpretation problem in the Copenhagen interpretation, and it can be thought of as the mind-body problem in some other theory--- you have a bunch of numbers describing the physics and you need to turn them into observer experiences. It's just incredible to me that such a subtle philosophical issue becomes so bone-jarringly obvious in quantum mechanics that everyone had to start thinking about it just so that they could calculate some atomic spectra.Likebox (talk) 19:17, 7 August 2008 (UTC)

"The end result is |Saw UP> + 2 |saw DOWN>." How can an end result (a collapse) be expressed as a ket? A ket is a quantum state, no? David Spector (talk) 12:27, 29 November 2011 (UTC)

Wigner's friend/quantum superposition/collapsing the wave function/Bell's theorem
Quantum physics/philosophy seems to have gotten itself tied up in knots by believing that what is in essence a solipsistic set of ideas, propounded by a forceful personality (Niels Bohr) and his heirs, is the framework by which we must consider the Einstein-Bohr debates, Schrodinger's cat, Wigner's friend, etc. At some point, one would think that more thinkers would back off and see that this all seems to be akin to medieval theology: How many angels can dance on the head of a pin?--can God make a knot that he can't untie? Does anybody find it Twilight Zone-ish to hear Wigner (and, earlier, von Neumann, I believe) speak of consciousness--which, it could be argued, is no more than a term referring solely to complex neural (i.e., physical) processes--as the cause of collapse? What level of consciousness? Is a mouse's sufficient for collapse, as Einstein sarcastically noted? What if humans can cause collapse, but on that day Wigner and his friend are too tired to make reliable observations, or their IQs have suddenly dropped? Under these circumstances, will their purported observatiions cause collapse? Of course, if one enjoys Bishop George Berkeley's philosophy, it is easy to see misled thinkers pondering as they do. (Does anyone read all of this and think of Mark Twain [I believe] who marveled at modern science as such a wonderful endeavor: One receives such a healthy return in conjecture for such a trifling investment of fact.)  For the growing number of thinkers who believe with Murray Gell-Mann (may we be forgiven for quoting him favorably) that Niels Bohr has brainwashed generations of physicists, I find that the book, Time's Arrow And Archimedes' Point by Huw Price, 1996, shows that, by seeing the truth of just one premise, these quantum conundrums vanish, and the realistic notions first pioneered by Schrodinger and Einstein become pre-eminent. In particular, I recommend pages 199-260 of the above cited book. There are other books by physicists: Victor Stenger's The Unconscious Quantum and John Gribbin's Schrodinger's Kittens And The Search For Reality, but I find Price's work to be the best. —Preceding unsigned comment added by 68.49.3.92 (talk) 20:39, 22 December 2008 (UTC)
 * That's why we have other interpretations of quantum mechanics. 76.67.72.232 (talk)

misunderstanding
Um, I showed my brother (a high energy physicist currently doing a PHD) this page, and he said that it showed why you should be careful what you consider a measurement, nothing more, and that it in no way proves consciousness. I don't expect you guys to accept that, but maybe some deeper research should be done and someone with more authority should look at this page.

As physics student I concur. This page says almost nothing about Wigner's Friend and what it says is rather misleading. Olleicua (talk) 05:34, 8 December 2009 (UTC)

I'm also a physics student and I agree. The article seems to have a strange understanding of what "measurement" actually means in quantum mechanics. 139.18.149.71 (talk) 12:59, 17 December 2011 (UTC)

OTHER INTERPRETATIONS
So.. there are a lot of interpretations of quantum physics. Many worlds has its own page. This page is about Wigner, not Everet. Why then is the majority of the page (and the only reference) devoted to Everet and Many worlds. This borders on an NPOV issue IMO because it makes it appear that there is only one major interpretation of QM. 216.158.163.188 (talk) 05:32, 8 December 2009 (UTC)

Sorry, the above comment is mine, I forgot to sign in. Olleicua (talk) 05:33, 8 December 2009 (UTC)

Hidden variable theories
I have the book cited above, Times's Arrow, and it says very clearly on page 200, "I think it is fair to say that the measurement problem is an unsolved one, despite heroic efforts."

Further, hidden variable theories, which are those that present the most credible alternative to the Copenhagen interpretation, have been cast in doubt by experiments that prove that Bell's inequalities do not hold (up to 242 standard deviations). See the wikipedia article Hidden variable theory. — Preceding unsigned comment added by Cliffberg (talk • contribs) 04:52, 2 February 2011 (UTC)


 * I think that claim is overly broad. Read the Hidden variable theory article more carefully and you'll see that Bell's theorem and Aspect's experiments argued only against theories in which all hidden variables are local, not against all theories of hidden variables. Stating this in Einstein's terms: God might not play dice but He sometimes does things spookily at a distance. SEppley (talk) 15:50, 27 November 2011 (UTC)


 * Okay, but those spooky things at a distance cannot transport information. Therefore, they must be illusions, just as much as stars appearing from Earth to move faster than c, or the point of closing of a scissors being able to move faster than c. That's an important point to remember. There is strangeness in QM, but nothing that results in miracles in the sense of violation of natural laws. QM has the potential to produce wonderful things (room temperature superconductors, room temperature superfluidity, and probably very fast quantum parallel computers), but not physical miracles like time travel of humans, time travel for human senses, time speedup or slowdown, antigravity devices, instant communication over long distances (even of one bit of information), perpetual motion, or teleportation of humans. David Spector (talk) 20:54, 18 December 2011 (UTC)

Seems like nonsense to me; what am I missing here?
What if there are N friends in the chain? Why would the collapse not happen at the first friend who observes? why should it happen at the 2nd in the chain, or the last in the chain? More to my reductio ad absurdum point, is it sensible to ask if a tree makes a sound if it falls in an unpopulated forest?

It seems to me that observation does not require consciousness, it just requires injecting energy into a system. Push in a little energy and you can observe the interference known as the Uncertainty Principle. Push in enough energy and you can generate new kinds of fundamental particles!

If observation (perturbing with energy then measuring the new state) collapses wave functions, then, by Occam's/Parsimony, we don't need to invoke consciousness. David Spector (talk) 02:12, 29 November 2011 (UTC)

"Experimental recreation"
I strongly object to this addition. The experiment has yet to be peer-reviewed, and moreover, the coverage is almost pure hype. We cannot build an encyclopedia upon unreliable sources. Nor is the description of the experiment accurate: the test was done on photons, making it not an example of the defining feature of the Wigner's-friend idea, which is placing an actual observer in a quantum superposition. In no way should the Proietti et al. paper be presented as a real implementation of Wigner's friend, or any extension thereof. Reading the paper beyond their grandiose abstract makes this absolutely clear. We cannot even argue that it is a notable un-peer-reviewed experiment, since the coverage derives entirely by cloning from a single, awful blog post at Technology Review. Echoing the blithering of the hype machine is not our job here. The text in question is simply unacceptable. XOR&#39;easter (talk) 20:30, 16 March 2019 (UTC)
 * And even if you take the (ludicrous) step of regarding a photon as an "observer", the experiment didn't implement Wigner's Friend, but a variation of it due to Časlav Brukner, so the text can't even get that much right. The sentence Wigner's friend experiment has been experimentally recreated, with the outcome as predicted by Wigner is wrong on both points, since it wasn't Wigner's proposal, and so the outcomes couldn't have been what Wigner predicted. With half the text being wrong, and the other half an admission the work is not yet published in a reliable source, we can't make the case for including it. The edit summary compares this case to that of FTL neutrinos, but there is really no analogy to be made. That was a report by a major collaboration which immediately attracted not just fawning press reports, but careful investigation by actual physicists (together with all kinds of speculation in the range between). This ... isn't that. Maybe it will be eventually, but Wikipedia is neither a rumor mill nor a prophecy service. We lose nothing by waiting for the formal cycle of physicist (and, in this case, potentially also philosopher) commentary and critique. XOR&#39;easter (talk) 21:34, 16 March 2019 (UTC)
 * I second everything XOR&#39;easter says above. This is nothing like FTL neutrinos and the experiment in question is a far cry from any sort of experimental verification of the Wigner's friend scenario. --Patchworld (talk) 00:08, 17 March 2019 (UTC)
 * Calm down. Yes, it is not notable (but notability applies to article topics, not to their content), it is not peer-reviewed, and it is a primary source.  I do not think that your objection that photons are not observers is sensible in the context.  In all, I think that a mention of this does not belong in the lead, but that it does merit a mention alongside the equally shaky  (which seems like it could do with severe trimming).  The article is not at the quality level where every mention of non-notable research must be excluded.  I would suggest that both these should be reduced to subsections of a section "Published papers", or suchlike.  —Quondum 01:10, 17 March 2019 (UTC)
 * I agree that the long section on extensions could stand trimming. It is rather, well, over-extended. In addition, it focuses excessively upon a single version of the idea, and not necessarily the first or the clearest. The first was, I believe, the version by Brukner, originally posted in 2015. The variant by Jeffrey Bub from 2018 might be less notationally demanding than what the article is currently trying to explain. (My use of the word "notable" merely echoed the edit summary to which I was responding. If I were in the mood to be a stickler for the peculiarities of Wikipedia's internal jargon, which I am sometimes, I'd call its inclusion in the lead a "violation of WP:UNDUE" instead.) Cheers, XOR&#39;easter (talk) 03:58, 17 March 2019 (UTC)

Contradictory addition
This edit introduces an error. The point of the "consciousness causes collapse" dodge is to ensure that the so-called paradox of inconsistent state assignments does not arise, because the friend is a conscious being so Wigner would not treat him as an ordinary quantum-mechanical system. It is not that there is a contradiction, specifically in the "consciousness causes collapse" interpretation, but that consciousness causing "collapse" (however that is supposed to work) escapes the contradiction. The contradiction is not specific to the "consciousness causes collapse" interpretation.

The article is already fantastically unclear; let's not make it worse, please. XOR&#39;easter (talk) 01:28, 18 November 2019 (UTC)
 * Fun fact maybe worth noting: at least in his later years, Wigner himself didn't actually hold to the "consciousness causes collapse" idea. He regarded that as the logical conclusion of the premises he had originally set out, but instead found the conclusion sufficiently absurd to be grounds to question the validity of those premises . XOR&#39;easter (talk) 01:38, 18 November 2019 (UTC)
 * Reading that whole section from the top again, it needs more thorough revision to tell apart the wavefunction-collapse-is-change-of-information way of thinking (Peierls, Zeilinger, Fuchs, Mermin, etc.) with wavefunction-collapse-is-a-physical-event-caused-by-consciousness (like in the Greg Egan novel Quarantine). XOR&#39;easter (talk) 22:09, 18 November 2019 (UTC)

recent experiment
Paper summary might be usable in article. 67.164.113.165 (talk) 08:39, 18 November 2019 (UTC)
 * Thank you for those links. The summary is by the authors themselves, which is both good and bad: it presents what they think without an intermediary, but it's not an evaluation by a secondary source. The copy at phys.org is a reprint of the original posting at The Conversation, here. XOR&#39;easter (talk) 17:57, 18 November 2019 (UTC)

Wigner's friend has nothing to do with Schrodinger's cat
Have none of the contributors of this article actually read what Wigner wrote? He doesn't once mention anything about cats, Schrodinger's or otherwise.

Look, here is Wigner's article: http://www.informationphilosopher.com/solutions/scientists/wigner/Wigner_Remarks.pdf

This article needs a total rewrite from somebody who knows what they're talking about. I don't know where all this talk of "dead cat/sad friend" came from, but it wasn't Wigner.

This article is a total shambles. Embarrassing.

No such thing as friend or amusing or not, laugh any. — Preceding unsigned comment added by Viwkls (talk • contribs) 01:04, 15 December 2019 (UTC)

From Wigners article: "One can either say that the wave function was invented to yield the proper probabilities, or that the law given in quotation marks has been obtained by having carried out a calculation with the wave functions, the use of which we have learned from Schrodinger. ", page 175. Before he describes a likewise experiment as with a cat, but with light flashes.

Also on : "Let's use the cat example, because it is more vivid. You can see Wigner's original argument on the Wigner's Friend page.. The physicist friend inside the lab opens the box and observes either a live or dead cat. But Wigner is outside the lab and does not know the outcome. Wigner says this seems to leave the world in a superposition of states - "dead cat/sad friend" and "live cat/happy friend." Wigner says that any inanimate material measuring device is left in a superposition of states. This would include his friend and himself, but for human consciousness. He resolves his paradox by saying that consciousness collapses the wave function, both his friend's inside the laboratory and his own." GerGroeneveld (talk) 04:21, 27 August 2020 (UTC)

News article in Science
"Paradox puts objectivity on shaky footing" George Musser  Science 369 Pages 889-890  28 August 2020 Charles Juvon (talk) 22:16, 30 August 2020 (UTC)
 * ✅ - Already considered and added - see this edit and related - iac - Drbogdan (talk) 22:25, 30 August 2020 (UTC)

Reinstate Reference?
I would like to consider reinstating the reference on this page to: -

Łukaszyk, Szymon (2018-01-25). "A No-go Theorem for Superposed Actions (Making Schrödinger's Cat Quantum Nonlocal)". New Frontiers in Physical Science Research. 3: 137–151. arXiv:1801.08537. doi:10.9734/bpi/nfpsr/v3/17106D

The reference was removed in November 2022 by Tercer as the author was considered self-promoting and B P International as a predatory publisher. My understanding is that B P international is not predatory (but I am basing this on a browse of their site and consideration of the sorts of publications they produce), and that the reference could be allowed to stand. — Preceding unsigned comment added by Travelkev (talk • contribs) 11:57, 28 November 2022 (UTC)
 * It is a predatory publisher, as listed here under the name "Sciencedomain International". Also, the edit was clearly self-promotion, as the user who added it is the author of the paper. I'm very curious about how you even find out about the removal of this source, given that you haven't edited in 6 years, and never in physics articles. Tercer (talk) 17:33, 5 December 2022 (UTC)

Extended Wigner's friend arguments
The section "An extension of the Wigner's friend experiment" should be changed to "Extended Wigner's friend thought experiments" to provide a summary of various extended Wigner's friend no-go theorems developed in the past 7~8 years, not just the Frauchiger-Renner theorem. For example, the Local Friendliness theorem is also a significant development in quantum foundations based on Wigner's friend ). The content in the current section could be moved to a separate wiki page for the Frauchiger-Renner no-go theorem. Ggstddup (talk) 12:22, 6 August 2023 (UTC)

Wigner's friend as a quantum circuit
This article includes the following: "On the other hand, one presentation of the experiment as a quantum circuit models the agents as single qubits and their reasoning as simple conditional operations. [21]" where the citation is the following blog post, which itself appears to be blank: Watching the Watchmen: Demystifying the Frauchiger-Renner Experiment — musings from Lidia del Rio and more at the 6th FQXi Meeting

I looked through Lidia de Rio's publication list on Google scholar and found this paper: Thought experiments in a quantum computer, which introduces a software library for running Wigner's friend as a quantum circuit. I'm not certain if this is the first article to propose a formulation of this thought experiment as a quantum circuit, but it seems a more appropriate citation for the above quotation. I reached out to the author to confirm this, but haven't heard back. PlantainPhysics (talk) 17:35, 26 March 2024 (UTC)


 * I would just remove that sentence. Modelling it as a quantum circuit is trivial and achieves nothing in the way of enlightenment. Tercer (talk) 20:55, 26 March 2024 (UTC)
 * It may be trivial for experts, but it seems appropriate to include in an introductory article like this. Enlightenment is in the eye of the beholder, and personally speaking, I found it interesting that this formulation had been implemented. I accept that the above citation was likely not the first time it was introduced. Perhaps given that David Deutsch extended the Wigner's friend thought experiment in as well as having a canonical quantum algorithm named after him, he may have connected these two at one point. I simply haven't read all the relevant papers from that era. 2601:602:8D00:3C70:CCCA:7D82:7ED:1BE9 (talk) 16:12, 27 March 2024 (UTC)