Wheeler's delayed choice doesn't change the past

[Alfredo Tifi]

I'm delightfully reading the Grand Design by Hawking-Mlodinow. After many years, I became persuaded from it that Feynman's interpretation as multiple paths of interference patterns in the double slit experiment and, more generally, his path integral approach to quantum mechanics, are superior to other mathematical settings because they capture some element of reality of "unobserved motion & causality" better than fickle behaviors interchanging between particle and wave. In other words, what is strange is the nature of motion or causal link between two consecutive observations-sources of information, not the fact that what is moving is a countable, detectable unit or "solid" and permanent object as a buckyball with a precise structure C₆₀.
What I'm going to claim and underscore here is that multiple (and simultaneous) stories in Feynman's calculation technique are not "true" stories in time; not in the sense of causative chain or sequence of events in time. I think everybody, included Feynman, was aware of this tenet. Multiple paths "capture" something that is inherent to the reality of motion, as a sniffling of proximal paths which explains randomness and relation between subsequent observations, but we don't know exactly what is it. Wheeler delayed choice experiment (WDCE)changes the future only (in the intercepted particles, not the others), does not affect the "true" past (in which the interference "was" actually generated) in those cases in which we are authorized to consider that as a real past event (in the cosmological thought experiment).
We must distinguish two cases. First, the real WDCE experiment in which we (erroneously) deduce a time delay from the spatial location in which we choose to observe/don't observe each particle. The Feynman's multiple paths do not correspond to actual single stories. So it is a nonsense to affirm that if the site of observation-choice is "subsequent" (actually to the right) of the two (open) slits, then it follows we are observing something which is occurring "after" in the Feynman's paths. We are actually making an observation set which is alternative to the observation on the screen. We can thus compare two alternative experiments, but not to consider one experiment as the continuation of the other. In conclusion, there is no difference between the choice to observe some of the particles just in front of the slits, behind the slits or close to the detector. In all these cases we are forcing the "possible" (Feynman's) stories of the observed particles with our present action and changing the future true story of those observed particles. But we can't say anything about the unobserved part of the single particle history (in between the source and the decision probe or the screen-detector).
Second, if we have a quasar light arriving at our planet after some billions of years, forming an interference due to the passage of photons through the two right and left ends of a big lens-galaxy, and if today the quasar and the galaxy are not anymore aligned nor existing, then each single photon arriving upon our interference screen today does have its true story. There is a true past: the emission from the quasar. Then we have only simultaneous possible paths passages through one possible side or the other of the lens galaxy and gravitational deflection & interaction with any other possible path. Eventually we have two different possible "present times" in the true history of the photon, depending on our decision: one in which the photon has been intercepted by the probe spaceship that was sent midway towards the galaxy, so that that photon won't go to form the interference pattern in case it will reach our planet (we have changed the future of that photon). The other "present time" is in case we are choosing not to observe the photon stream for a while. After that the unobserved photons conserve the interference path that was generated during the past history and will arrive on the interference screen on Earth, plotting the interference pattern. In this case, the true story ends with the impact of the photon on the screen-detector on Earth. Our decision not-to-observe has not changed in any way the true story of the single photon.
They are not our "decisions" which change the history of events, but only facts, actual events can do that: if we actually interact with-observe the single photon, that interaction will determine a really new future chain of events for that photon. There is no way to make any interaction - or decision that won't cause a true interaction - which would determine a change in some true past event of a true history, as the emission of photon from the quasar or as the same existence of a certain photon we haven't observed yet, with or without interference.

 

[vanhees71]

Well, Feynman's path-integral formulation of QT is equivalent to the uniquely defined QT. There's no more qualitative insights than with any other formulation of QT (including relativistic QFT, where the correct path integral is over field configurations, not paths of particles in phase space as is possible in non-relativistic QT).

Of course, the delayed choice is not changing the past but it's rather postselecting which aspect of the situation you want to observe. Last week, I defended my habilitation about Walborn et al's realization of Wheelers ideas, which I find most intuitive. In the discussion, one of the professors stated: "But the photons is already absorbed! How can one then "postselect" whether to observe wave-like or particle-like properties after this?" He was satisfied by my answer that you can do the postselection due to the preparation of the photon pair in the polarization-entangled pair (Bell state). Here are my slides (although in German, maybe you can understand it):

https://th.physik.uni-frankfurt.de/~hees/publ/habil-coll-talk.pdf

 

[Demystifier]

Delayed choice, indeed, does not change the past, as I explain concisely in https://arxiv.org/abs/1703.08341 Sec. 5.1.

 

[Alfredo Tifi]

Thank you Hendrick and Hrvoje. I couldn't fully understand your papers either they were both in Italian or English, but - after your replies - I believe Hawking is exaggerating in telling us (page 83):
«...the Universe doesn't have just a single history, but every possible history, each with its own probability; and our observations of the current state affect its past and determine the different histories of the Universe, just as the observations of the particles in the double slit experiment affect the particles' past.»
Do you agree?

 

[vanhees71]

Hm, the only book by Hawking I like is Hawking, Ellis, The Large Scale Structure of Spacetime...

 

[Lord Jestocost]

«...the Universe doesn't have just a single history, but every possible history, each with its own probability; and our observations of the current state affect its past and determine the different histories of the Universe, just as the observations of the particles in the double slit experiment affect the particles' past.»

I am astonished, that this has been written by Hawking.

Regarding delayed choice experiments: To my mind, Wheeler merely suggests that it doesn't make sense to talk of "reality" before a measurement is made:

„The "past" is theory. The past has no existence except as it is recorded in the present. By deciding what questions our quantum registering equipment shall put in the present we have an undeniable choice in what we have the right to say about the past.“ [Underline, LJ]

Wheeler in „Quantum Theory and Measurement“ (edited by John Archibald Wheeler and Wojciech Hubert Zurek), Princeton, New Jersey 1983, page 194

 

[Alfredo Tifi]

Here is the book, see pages 82-83 https://goo.gl/images/Z6N9gr (maybe Mlodinow wrote that).

 

[Alfredo Tifi]

I am astonished, that this has been written by Hawking.

Maybe Mlodinow wrote that. Here is the book, check pages 82-83 https://goo.gl/images/Z6N9gr.

 

[DrChinese]

... I believe Hawking is exaggerating in telling us (page 83):
«...the Universe doesn't have just a single history, but every possible history, each with its own probability; and our observations of the current state affect its past and determine the different histories of the Universe, just as the observations of the particles in the double slit experiment affect the particles' past.»
Do you agree?

Naturally, almost anyone would find something to quibble with in Hawking's statement. After all, it is sweeping in the assumptions inherent in it.

For example: I don't agree that the Universe has all possible histories, nor all possible futures. The most it could contain are histories and futures consistent with what we observe.

On the other hand: there is no evidence that otherwise refutes Hawking's statement. And one possible deduction is that past (and future) histories are "pruned" as a result of measurement decisions made now. That implies a weak form of retrocausation. I say "weak" because you are not selecting/pruning specific histories with a measurement choice.

 

[Good Elf]

Of course, the delayed choice is not changing the past but it's rather postselecting which aspect of the situation you want to observe.

Good point! There is no strict necessity in Quantum Theory for the Uniqueness of quantum phenomena, it is only that separate individual observers will not see mutually inconsistent outcomes from the observed events that other observers may see. Phenomena exist depending on who and how we observe them, in fact the observation call them into existence by "fiat", so there may be more than one self consistent outcome which may turn out to be mutually inconsistent from the POV of another observer, so there may be more than one Everett World in a series of overlapping event spaces like a number of criss-crossing roads along which a single observer may "travel". The "roads" may be traveled many times and in many ways... a form of event based eternalism based on a Page & Wootters Mechanism for time in a Two-State Vector Formulation. We as observers don't notice it because we are causally driven creatures.

I am partial to the Many Interacting Worlds Interpretation of Hall, Deckert and Wiseman. We live in a Block Universe. The superposition of all quantum states (a closed set of a large number of mutually overlapping results, but not an infinite number) is an influence in this world from the "quantum overlap" of "classical" parallel Universes. A specific measurement "selects" from a background of many outcomes which local bit of the Universe will happen as a series of following consistent classical states. Each Universe is classical and bears a resemblance to our own Universe with differences, but the many quantum states in superposition are individually real and legitimate alternate choices.
A New Hypothesis Suggests That Parallel Universes Might Interact after All - Big Think - December 2, 2017
or the paper...
Quantum Phenomena Modeled by Interactions between Many Classical Worlds - Michael J. W. Hall, Dirk-André Deckert, and Howard M. Wiseman - Phys. Rev. X 4, 041013 – Published 23 October 2014
Still... a fruitful developing idea in recent developments (2017). It is version consistent with some of the more interesting counterfactual experimental outcomes that can be actually observed.

 

[vanhees71]

After reading Hawking's "Short History of Time" concerning "imaginary time" I was also shocked. As I said, the only book (co-)authored by him I like is the Large Scale Structure of Spacetime. It's of course much more difficult to write popular-science books that are correct and at the same time understandable to the addressed non-expert audience than to write a good textbook addressed to students of theoretical physics who are supposed to have the necessary technical prerequisites to understand the subject. That's the only excuse I have for scientists writing misleadin popular-science books. That you can do better has been shown by the (in fact very few) examples of good popular-science books (Weinberg, The first Three Minutes; Ledermann+Teresi, The God Particle (title given by a stupid publisher not by Lederman himself, but that's the only drawback of this one); Born, Einstein's Theory of Relativity (which in the 1920ies helped to keep Frankfurt university's physics department running)).

 

[Good Elf]

Seems there is some interest in this topic. So I would like to add a comment or two to what I have already stated above regarding the "Block Universe". Assuming the Universe is indeed a "Block Universe" and time is related to quantum state entanglement via an experimentally demonstrated Page & Wootters Mechanism indicated by the experimental quantum entanglement technique referenced in this paper: Time from quantum entanglement_an experimental illustration - LMaccone&MGenovese - 1710-2013
A Two-State Vector Formulation, as espoused by Vaidman/Feynman/Kastner/Cramer and others, would allow a very flexible view of time, almost "mechanical" in content, and leads to the possibility of a local reversal of events at the quantum state level... virtually unscrambling the egg (as it were).
More recently this article in Science Alert and also in Nature News have these arguable examples which are highly supportive of "an Eternal Block Universe" which might have the network of quantum events being repeatedly traversed through a "reusable" criss-crossed network of interconnected Everett Worlds, yet individually reversible events (at least at the quantum electrodynamic level): Physicists Have Created a Set of Conditions in Which Time Seems to Run in Reverse - Science Alert - 2712-2017, and The new thermodynamics: how quantum physics is bending the rules - Nature News - 0111-2017 and these experiments seems to be just the prelude to the new paper:

Reversing the thermodynamic arrow of time using quantum correlations - Kaonan Micadei etal -0911-2017.
The second law permits the prediction of the direction of natural processes, thus defining a thermodynamic arrow of time. However, standard thermodynamics presupposes the absence of initial correlations between interacting systems. We here experimentally demonstrate the reversal of the arrow of time for two initially quantum correlated spins-1/2, prepared in local thermal states at different temperatures, employing a Nuclear Magnetic Resonance setup. We observe a spontaneous heat flow from the cold to the hot system. This process is enabled by a trade off between correlations and entropy that we quantify with information-theoretical quantities.

In no way is this "time travel" but phenomena occurring in time might be retrospectively changed applying the Wheeler Quantum Eraser Paradigm, if some quantum event cause some disastrous following chain of events in a single Everett World, it may be possible through the exercise of a strong guiding quantum correlation to "dodge the quantum bullet". The fact that we are using only "strong quantum correlations" rather than even stronger quantum entanglement, it might be possible to accomplish an outcome by "postselecting which aspect of the situation you want to observe" as Vanhees71 suggested, an aspect different from the first erroneous and disasterous sequence of events, by first invoking the NMR rollback technique followed by rolling forward and correlating an alternative classical outcome through a forcing "observation", or by simply preventing the incorrect "observation" happening by "editing it from this temporal script". It is science fiction ... but it is not speculation and is based on desktop experiment. A quantum Schrodinger's Cat could be brought back from the dead if we wanted to if we accepted the Many Interacting World's Interpretation of Hall, Deckert & Wiseman. Or Gort can resurrect Klaatu, breaking the seeming law of the irreversibly of death. I imagine there will be a practical time limit involved where such chain of event reversibility would become practically possible. But for now, we don't know what we don't know. Any other ideas out there? Doctor Strange eat your heart out.

 

[StevieTNZ]

„The "past" is theory. The past has no existence except as it is recorded in the present. By deciding what questions our quantum registering equipment shall put in the present we have an undeniable choice in what we have the right to say about the past.“ [Underline, LJ]

Wheeler in „Quantum Theory and Measurement“ (edited by John Archibald Wheeler and Wojciech Hubert Zurek), Princeton, New Jersey 1983, page 194

I am of the same opinion. It is an unobserved past that gets defined into a definite past, when the present measurement takes place.

 

[Demystifier]

I am astonished, that this has been written by Hawking.

Hawking wrote a lot of nonsense in last 10 years or so.

 

[entropy1]

If there is retrocausality, then the fact of object A retrocausing event X is fixed from the occurence of event X up to the retrocausation by A taking place. That is, when X occurs, the retrocausation by A is fixed. But I see it a little differently. Let's call the retrocausation by A "Y". So Y retrocauses X. Then X causes Y. So we have a causal loop. Put differently: neither of the events X or Y is causing each other, however their outcomes depend on each other.

That is my opinion; I will not motivate it for I am not qualified enough and it would be speculation.

 

[vanhees71]

Well, within the established and well-working theories of physics, among them local, microcausal relativistic QFT (leading to the Standard Model that up to now withstands all hard work to disprove it with bravour) there is no retrocausation possible, and there's not a single observation today that hints at its existence at all. What is an established fact are indeed the correlations described by quantum entanglement, i.e., correlations that are stronger than any that can be described by local deterministic hidden-variable theories.

 

[entropy1]

I think you are adressing me, @vanhees71 : do you mean that correlations with the past are possible? In my eyes correlation could be formulated in terms of two-way causality, which is something different than one-way causality.

I fully respect QT as a correct theory, and if I am right, what I claim should be part of QT and a result of it. Probably I am saying something that, in other words, has been addressed a long time ago by people who are knowledgeable and competent to do so (like you! ).

 

[vanhees71]

I don't know what you mean by "correlations with the past", but of course the correlations, described by entangled states, are due to preparation of the system under investigation in the past, e.g., the nowadays widely used polarization-entangled photon pairs created by parametric downconversion: The Bell-inequality violating correlations are due to the preparation of the photon pair before any measurement is done with them, and this correlation also enables the postselection of "wave properties" or "particle properties" as described in quantum-eraser experiments (which are one particularly nice example for the possibility of Wheeler's delayed-choice idea). With a detailed enough measurement protocol you can do this postselection long after the entire experiment, including all photons, is long gone, and there is for sure no mystic retrocausation involved. It's just selecting subensembles of measured sequences of photon detections.

 

[Lord Jestocost]

I am of the same opinion. It is an unobserved past that gets defined into a definite past, when the present measurement takes place.

I, personally, would render Wheeler’s statement more precisely and would say: By deciding what questions our quantum registering equipment shall put in the present we have an undeniable choice in what we have the right to say about the past in classical terms.

 

[DrChinese]

... there is no retrocausation possible, and there's not a single observation today that hints at its existence at all...

That's not a fair statement. It is more fair to say that such experiments can be interpreted in other ways. For example: the below referenced experiment is dismissed by you (as we have discussed previously). However, I present to thread readers an example (from a top team) of entangling particles via swapping. The swapping occurs AFTER the entangled particles have already been detected. That is certainly an observation that hints at retrocausation.

https://arxiv.org/pdf/quant-ph/0201134.pdf
Middle of page 5 is the entangle-swap after detection: "... Alice’s measurement projects photons 0 and 3 into an entangled state after they have been measured."

Of course, there is no signalling possible with this scheme, as the projection of the specific entangled state is itself random.

 

[Good Elf]

Correlation is not the same as causation. Events have causes and they have effects while events have correlations. In the Quantum Universe (which affects all physics, including classical physics) there are ways to create quantum correlations. They are like flies on pigs... these correlations exist even if we are not "personally" tracking them. The current techniques usually use a pair of down converted photons, but correlations are stronger than that, and other means may be found to track them in the future. The Classical Universe, the one "we" are living in, is a continuous emergent process, emergent in time, that is hidden from view when we make a classical measurement (an observation). We can track the past but we cannot know the future. Classical observations do not reveal this otherwise hidden entanglement/correlation. It is possible to understand correlations in the form of quantum entanglement which is a form of true instantaneous action at a distance. The influence of Classical Measurements are always bound within the light cone, while entanglement can act in spaces beyond the bounds of the classical light cone, and a systems correlation from beyond at that level could theoretically still interact with the events inside the expanding light cone and potentially reverse them. Nobody knows how it works but we know it does work since we can run specifically prepared experiments that show truth in these approaches.

Classically speaking there appears to be no retro-causation. On the other hand there are many quantum experiments that show that in the quantum universe of the superposition of events, this superposition of states may be interpreted as the result in our Universe of interactions between all the available Classical Everett Worlds, which is not infinite but may still be a very large number. We can choose an outcome that forces a specific correlation. Further... the correlation of events that are strongly correlated can be run in either direction. All the "laws" of quantum physics are time symmetric and indicate that quantum electrodynamics, is similarly symmetric with respect to time. Of course Feynman's PhD Thesis was exactly this and led eventually to his Nobel Prize winning version of Quantum Electrodynamics. However he confided in his biographer before his death that he was always very fond of his "Absorber Theory", but could not prove it's truth in his lifetime. So far we have no direct indication that other forces are correlated in this same way as Electromagnetism, but the Strong and Weak Forces are range limited, meaning correlations beyond the range of those forces may affect those processes acausally in a similar way to the way events in quantum electrodynamics might be influenced from beyond the light cone. Naturally it is not expected that even running the event backwards in time will work entirely if some effects of Classical Change is already evident to the external observer or escape the provisions of other range limited forces. A correlation induced reversal will have less of a global classical influence beyond that outer limit in time or in space. An obvious Quantum Mechanics example is the Quantum Zeno Effect, an article that discusses these influences is here:
Why the Watched Pot May Actually Never Boil, According to Quantum Physics - Epoch Times - 29 June 2014. In the quantum world, a watched kettle never boils as continuous entangling measurement resets the quantum state back to the initial state. This effect may be used to prevent an unstable atomic nucleus from decaying from the POV of "our" emergent Classical Universe. Also judicious timing of the "observations" of the unstable state may accelerate it's decay, called the Quantum Anti-Zeno Effect. All of these theoretically predicted phenomena are also experimentally verified. Additionally the effects of unobserved quantum states may still be tracked/betrayed by what is happening in the surrounding regions as the effects of previous events spread in space and time. You can't stop other external phenomena from intervening. Like observing the wake left by a moving boat in the water. You don't need to see the boat to know it is there.

The only "clue" we have to this preference for direction is it's entropy. From the above experiments we know there is no inevitability in our Classical World, even the effects of thermodynamics which superficially appear in the classical world to be purely random chance, but Quantum Effects can be deterministically called up to change emergent outcomes. In this most recent experiment correlated quantum events may be run in either direction of time (see above)... leading "back" to a reset on the entropy of certain states... a system that has cooled down due to the process of classical thermodynamic and "entropy", if it is correlated strongly this process can be made to run "uphill" resetting/refreshing the state to a former condition when it was still "hot". No heat is added... it returns to it's former state through invoking correlations. An analog of this NMR "Echo" retrieval process is shown in this simple classical lab experiment.

This may not be "time travel", it is obviously something else in our "Block Universe" of interconnected criss-crossing, temporally static, Page & Wootten related events, it is handled by Quantum Information Theory and the still belongs to the theoretical Beckenstein Bound of "Black Hole Theory", only some of which I personally accept at this time. Inside our Universe the Light Cone defines the speed of electromagnetic influence. There would be a zone of quantum correlation for our entire "Black Hole Universe" existing beyond the "Black Hole" of our current Everett Universe and existing at a finite range "out there", no light signals from inside our Black Hole will ever reach beyond our Event Horizon. The inside boundaries of our Universe are also finite. However entanglement from beyond our Classical Universe does exist out there, and knows no such boundary... at least theoretically anyway. So from that one "external" perspective at least, all information content of our Universe "exists" eternally... that include all the history of time and space since the "Big Bang" (whatever that might really be), from the action of sub-atomic particles to the collisions of "Black Holes" and to the events that brought everything into being, is forever "frozen" into that Beckenstein Bound. At that level our "external" Classical Universe, a "Black Hole" is a Quantum Computer/Simulator that can order up any possibility by changing the boundary conditions and then "play them all" in any direction of time in any volume of space, a process we will probably proving is not only possible, but incredibly useful and technically available to us. The cosmic joker would say.. "We can run but we can't hide from that".

 

[DrChinese]

Seems there is some interest in this topic. So I would like to add a comment or two to what I have already stated above regarding the "Block Universe". ...

There is a recent interpretation of QM called Relational Blockworld. It is now quite well developed. It is an adynamical/acausal theory (they do not consider it retrocausal). You might be interested in it:

Relational Blockworld: Towards a Discrete Graph Theoretic Foundation of Quantum Mechanics
https://arxiv.org/abs/0903.2642
We propose a discrete path integral formalism over graphs fundamental to quantum mechanics (QM) based on our interpretation of QM called Relational Blockworld (RBW). In our approach, the transition amplitude is not viewed as a sum over all field configurations, but is a mathematical machine for measuring the symmetry of the discrete differential operator and source vector of the discrete action. Therefore, we restrict the path integral to the row space of the discrete differential operator, which also contains the discrete source vector, in order to avoid singularities. In this fashion we obtain the two-source transition amplitude over a "ladder" graph with N vertices. We interpret this solution in the context of the twin-slit experiment.

One of the authors is a PF member. They have a book that just came out which goes much deeper:
https://global.oup.com/academic/product/beyond-the-dynamical-universe-9780198807087?cc=us&lang=en&

 

[Alfredo Tifi]

I don't agree that the Universe has all possible histories, nor all possible futures. The most it could contain are histories and futures consistent with what we observe. ...
one possible deduction is that past (and future) histories are "pruned" as a result of measurement decisions made now. That implies a weak form of retrocausation. I say "weak" because you are not selecting/pruning specific histories with a measurement choice.

Maybe this kind of "pruning" coincides with Hawking top-down approach, and both remember to me something similar to an "Anthropic Principle".
I want to quote Hawking's argument about his application of Feynman sum to the entire universe, two chapters later (pages 135, 136, 140).

«If the origin of the universe was a quantum event, it should be accurately described by the Feynman Sum Over Histories. To apply quantum theory to the entire universe - where the observers are part of the system being observed - is tricky however. In Chapter 4 we saw how particles of matter fired it at a screen with two slits in it could exhibit interference patterns just as water waves do. Feynman showed that this arises because a particle doesn't have a unique history. That is, as it moves from its starting point A to some endpoint B, it doesn't take one definite path, but rather simultaneously takes every possible path connecting the two points. From this point of view, interference is no surprise because, for instance, the particle can travel through both slits at the same time and interfere with itself. Applied to the motion of a particle, Feynman’s method tells us that to calculate the probability of any particular endpoint we need to consider all the possible histories that the particle might follow from its starting point to that endpoint.»

I have underlined to highlight that Feynman multiple histories between starting point A and endpoint B don't need to be necessarily true histories. They can be just mathematical devices to calculate the probability of B given A. But, then, he writes:

«One can also use Feynman’s methods to calculate the quantum probabilities for observation of the universe. If they are applied to the universe as a whole, there is no point A, so we add up all the histories that satisfy the no-boundary condition and end at the universe we observe today.
In this view, the universe appeared spontaneously, starting off in every possible way. Most of these correspond to other universes. While some of those universes are similar to ours, most are very different. They aren’t just different in details … but rather they differ even in their apparent laws of nature. In fact, many universes exist with many different sets of physical laws. Some people make a great mystery of this idea, sometimes called the multiverse concept, but these are just different expressions of the Feynman Sum Over Histories.»

In this case, the underlying is to notice that the multiple histories by Hawking are becoming true histories, although in other universes. Finally, he writes:

«But there will be different histories for different possible states of the Universe at the present time. This leads to a radically different view of cosmology, and the relation between cause and effect. The histories that contribute to the Feynman's sum don't have an independent existence, but depend on what is being measured. We create history by our observation, rather than history creating us.»

The first underscore is to confirm the existence of the multiple histories in the physical world, in Hawking's vision. The second is a maybe stronger form of strong Anthropic Principle, in which it is our observation that determinates which one of the many parallel universes we have been inhabiting (without weakly or strongly pruning-retrocausating the others). So we are not "creating" our history, but "post-selecting" it.

 

[stevendaryl]

For example: I don't agree that the Universe has all possible histories, nor all possible futures. The most it could contain are histories and futures consistent with what we observe.

To me, the Copenhagenish belief that "nothing is real unless it is observed" seems inconsistent with the apparent fact of basis-independence of quantum mechanics. When it comes to something microscopic, such as whether an electron has spin-up or spin-down, we say that there is no fact of the matter unless it was measured. On the other hand, a measurement is simply an amplification of a microscopic property so that it produces a macroscopic effect. So the Copenhagen rule means that macroscopic properties are treated differently than microscopic properties. But the evolution equations don't make such a distinction.

I know that decoherence does naturally lead to a macroscopic/microscopic distinction, because in practice, interference effects between alternative macroscopic properties are unobservable. Since interference effects are the only clue we have that multiple possibilities coexist, the lack of macroscopic interference allows us to assume that only one alternative exists. But I would say that the theory itself doesn't really support that.

 

[Alfredo Tifi]

I like your wise reply and opinion about Copenhagen break up between micro/macro realities.
That means that either the theory is incomplete or classic reality is an illusion of the macroscopic world.
I know speculation is not liked in this forum, but it is evident that the classic idea of reality is based on the relationship and asymmetry between the history which "precedes" the measurement and the prevision of future event probabilities "after" the measurement. I.e. it is strongly based on a classical conception of time.
Many ideas about time, included the "M theory" as described by Hawking (time is what remains after the inflation and interferences among the eleven dimensions in our own universe, but it was in origin a spatial dimension as others) kill any residual hope to consider time classically in the future.
Furthermore, practically all the theories after QED seem to prefer the Sum over Histories scheme. SoHs are not yet physically meaningful, but we remember field lines from Faraday started as artificial crutches to become, eventually, more real than matter.
So, it seems to me very likely that the second option will come true: classic-macroscopic reality would result in just an illusion, and a different kind of microscopic probing will substitute current macroscopic measures in building a more reliable form of quantum mechanics and knowledge. In other words, what is limiting in the Copenhagen orthodoxy is the primate of the human-scale macroscopic observation, while we should recognize that it is a historical residue of the passage from classical to microworld (correspondence principle). For example, the true result of a "measurement" experiment should be deduced from nanoscale experiments in which the Planck constant or quantum noise are relevant. Obviously, deductions or interpretations would be more mingled with the theory on the bench than it occurred the past. Anyway, it is possible that in this new-non-orthodox quantum theory the multiple histories would acquire physical meaning.

 

[OCR]

I just cannot parse this statement at all... did you leave out some words, or punctuation, or is it just me ? .

Obviously, deductions or interpretations would be more mingled with the theory on the bench than it occurred the past.

 

[RUTA]

There is a recent interpretation of QM called Relational Blockworld. It is now quite well developed. It is an adynamical/acausal theory (they do not consider it retrocausal). You might be interested in it:

Relational Blockworld: Towards a Discrete Graph Theoretic Foundation of Quantum Mechanics
https://arxiv.org/abs/0903.2642
We propose a discrete path integral formalism over graphs fundamental to quantum mechanics (QM) based on our interpretation of QM called Relational Blockworld (RBW). In our approach, the transition amplitude is not viewed as a sum over all field configurations, but is a mathematical machine for measuring the symmetry of the discrete differential operator and source vector of the discrete action. Therefore, we restrict the path integral to the row space of the discrete differential operator, which also contains the discrete source vector, in order to avoid singularities. In this fashion we obtain the two-source transition amplitude over a "ladder" graph with N vertices. We interpret this solution in the context of the twin-slit experiment.

One of the authors is a PF member. They have a book that just came out which goes much deeper:
https://global.oup.com/academic/product/beyond-the-dynamical-universe-9780198807087?cc=us&lang=en&

Here is a more recent paper on RBW http://www.ijqf.org/wps/wp-content/uploads/2015/06/IJQF2015v1n3p2.pdf. There is a series of Insights on Blockworld (aka Block Universe) that I wrote which actually formed the starting point of our book. Here is the first in that series https://www.physicsforums.com/insig...time-dilation-length-contraction/#toggle-id-1. These Insights aren’t promoting RBW (rarely mentioned in fact), they’re just pointing out the power of adynamical explanation in the block universe where dynamical explanation in the mechanical universe leads to confusion.

 

[bhobba]

I am astonished, that this has been written by Hawking.

Of course - so am I - but to be fair in the transactional interpretation it just may. So the correct response is we don't know - but most would say - of course not.

Thanks
Bill

 

[vanhees71]

That's not a fair statement. It is more fair to say that such experiments can be interpreted in other ways. For example: the below referenced experiment is dismissed by you (as we have discussed previously). However, I present to thread readers an example (from a top team) of entangling particles via swapping. The swapping occurs AFTER the entangled particles have already been detected. That is certainly an observation that hints at retrocausation.

https://arxiv.org/pdf/quant-ph/0201134.pdf
Middle of page 5 is the entangle-swap after detection: "... Alice’s measurement projects photons 0 and 3 into an entangled state after they have been measured."

Of course, there is no signalling possible with this scheme, as the projection of the specific entangled state is itself random.

Entanglement swapping or teleportation is no hint at retrocausation either, because to enable it again you need also well prepared entangled photons, which then by coincidence measurements let's you, also in a delayed-choice setup, entangle other pairs which never have been in causal contact to an event before. Again, there cannot be retrocausation in QED (and any local microcausal relativistic QFT as is underlying the Standard Model), and as far as I know, nobody ever claimed that entanglement swapping, Bell experiments, etc. contradict standard QED/relativistic QFT, and indeed one can describe all these experiments within the standard theory. So there is no need to assume retrocausation and other acausalities based on any empirical finding yet, and as long as this is the case, there is no reason to give up the scientific method, which is strongly based on causality. Acausal behavior would be the end of science as we know it!

 

[vanhees71]

There is a recent interpretation of QM called Relational Blockworld. It is now quite well developed. It is an adynamical/acausal theory (they do not consider it retrocausal). You might be interested in it:

Relational Blockworld: Towards a Discrete Graph Theoretic Foundation of Quantum Mechanics
https://arxiv.org/abs/0903.2642
We propose a discrete path integral formalism over graphs fundamental to quantum mechanics (QM) based on our interpretation of QM called Relational Blockworld (RBW). In our approach, the transition amplitude is not viewed as a sum over all field configurations, but is a mathematical machine for measuring the symmetry of the discrete differential operator and source vector of the discrete action. Therefore, we restrict the path integral to the row space of the discrete differential operator, which also contains the discrete source vector, in order to avoid singularities. In this fashion we obtain the two-source transition amplitude over a "ladder" graph with N vertices. We interpret this solution in the context of the twin-slit experiment.

One of the authors is a PF member. They have a book that just came out which goes much deeper:
https://global.oup.com/academic/product/beyond-the-dynamical-universe-9780198807087?cc=us&lang=en&

Is this still science? Is this something beyond standard QT or just another mathematical technique to evaluate path integrals? I'm a bit sceptical ;-)). Last but not least the preprint (on a theoretical subject) is written in Word...

 

[bhobba]

On the other hand, a measurement is simply an amplification of a microscopic property so that it produces a macroscopic effect.

I think I get your drift in that decoherence transforms a superposition to a mixed state - but 'amplification' is not the word I would use - more like something along the lines of making 'clearer' by interacting with it. Also what it interacts with determines exactly what is now 'clearer'. Sorry for the semantic quibble - I usually think semantics is a rather silly thing to ague about, but sometimes it does help in clarifying things for others.

Thanks
Bill

 

[Lord Jestocost]

To me, the Copenhagenish belief that "nothing is real unless it is observed..."

That means that either the theory is incomplete or classic reality is an illusion of the macroscopic world.

To my mind, Copenhagen says nothing but: Microscopic entities are not „real“ in the sense that we cannot speak about them directly; we cannot describe these entities at all, we can only represent them by means of a quantum mechanical formalism; as distinguished from macroscopic entities which allow at least an approximate classical description.

 

[bhobba]

To my mind, Copenhagen says nothing but: Microscopic entities are not „real“ in the sense that we cannot speak about them directly; we cannot describe these entities at all, we can only represent them by means of a quantum mechanical formalism; as distinguished from macroscopic entities which allow at least an approximate classical description.

Basically yes - but there are issues, just like with ordinary probability what exactly does probability mean eg frequentest vs Bayesian. Copenhagen is more Bayesian - Ensemble more frequentest. And then their is the issue of is it complete.

Your view is more like that of Dirac. Heisenberg and other Copenhagenists thought it was complete - Einstein thought not - Dirac was not sure of either - he was open:
http://philsci-archive.pitt.edu/1614/1/Open_or_Closed-preprint.pdf

That's part of the reason I think of all the early pioneers Dirac was the closest to correct.

Added Later:
Oh I forgot to mention future developments seem to prove him correct - QFT was not a big change from the principles of QM but was certainly closer to 'correct' accounting for things that standard QM could not account such as spontaneous emission.

Thanks
Bill

 

[stevendaryl]

Basically yes - but there are issues, just like with ordinary probability what exactly does probability mean eg frequentest vs Bayesian. Copenhagen is more Bayesian - Ensemble more frequentest. And then their is the issue of is it complete.

Your view is more like that of Dirac. Heisenberg and other Copenhagenists thought it was complete - Einstein thought not - Dirac was not sure of either - he was open:
http://philsci-archive.pitt.edu/1614/1/Open_or_Closed-preprint.pdf

That's part of the reason I think of all the early pioneers Dirac was the closest to correct.

Thanks
Bill

I don't see how Bayesian versus Ensemble matters for the interpretation problems. It seems to me that the "rules of thumb" for using QM make a distinction between macroscopic and microscopic systems. Microscopic systems can be in superpositions, and so don't have definite values for observables, in general. Macroscopic systems, particularly measuring devices, are assumed to give definitive answers. In the terms you've used so often, the distinction is between improper and proper mixed states. QM doesn't describe how an improper mixed state becomes a proper mixed state. So if that actually happens, it seems that QM is incomplete, and the various interpretations of probability don't seem to make a difference (or I don't see how).

Of course, there is a possibility that the transition never happens, which to me implies Many Worlds, which has its own interpretation problems.

 

[bhobba]

I don't see how Bayesian versus Ensemble matters for the interpretation problems.

Well Copenhagen states the state is subjective, the ensemble states its a conceptual very large number of possible outcomes (actually infinite - but I don't know about you but I can't imagine that) associated with each observable. They are different - the same way Bayesian and frequentest are different. Is it of any practical importance - IMHO no. Having studied probability nobody that applies it worries about it, although sometimes they, without stating there is controversy about it, chose one in solving problems, but usually it doesn't matter at all. Same in QM - but here people like to make a distinction - in practice it makes just as much difference - basically none.

In the terms you've used so often, the distinction is between improper and proper mixed states. QM doesn't describe how an improper mixed state becomes a proper mixed state. So if that actually happens, it seems that QM is incomplete, and the various interpretations of probability don't seem to make a difference (or I don't see how).

You can look at it that way - Einstein probably would. If that's your won't - go for it - fine by me.

The other view is every theory, every single one has things that are simply accepted as true - that improper mixed states become proper ones is simply one that's accepted as true - you may find an explanation for that - and of course such would be an advance - but then you are faced with the question - why is that true? It has no logical status any different than the original explanation - aside from truth of course. The new explanation you may like, not like or even think it's worse. All this is just a personal reaction. I am in that camp - its just an assumption we make. It may have a deeper explanation - or not. Personally it doesn't worry me one way or the other.

These things are nothing but personal preference. I have mine - you have yours - Einstein has his. I don't really care. Any could be correct - doesn't bother me one way or the other. Got an experiment to decide - then progress has been made. Without experimental support the only benefit IMHO is they shed light on the formalism which is the important thing and why I like reading them. I don't care what one you or anyone else chooses - I just want to understand the formalism better.

Thnaks
Bill