If you wish
to put a letter in our Correspondence section (this one),
to signal us a relevant Publication or Link to be added in our lists,
to ask some questions about the Experiments and their implications,
please send us an e-mail.
Jump to cor 030301 cor 030302 cor 030303 cor 030304 cor040101
January 1, 2004
Regarding the article about your experiment in PM-Magazin (Germany), November 2003, I would like to put the following 3 questions:
Why does a photon once passe a half-silvered mirror and once not? If one could explain this behavior exactly, then maybe one could also explain the identical behavior of two different photons.
How are you producing the photons, which behave the same way at the mirrors? Do you separate them by another (half-silvered) mirror? If yes, it sounds logical, that they behave the same way. If they are produced consecutively by a (one) laser, it is not less logical for me that they will behave the same way.
Or maybe the universe is pulsing, so that all half-silvered mirrors in the universe will pass the photons for a while and later not, so that the behavior will be clear - even if there is a small time lack between the arrivals of both photons.
If they are building a relationship by accident, how do the photon, which will pass the mirror, know, that there is another photon, which will pass another (or the same) mirror in a few ticks? Will it call "Is there another photon anywhere in the universe which will pass a mirror? Please behave the same way" - I think not. Why else should they communicate with each other? I do more believe, that the production of the photons is the key to the same behavior, not the way, the target or something else.
Regarding the experiment with the moving mirror. I don’t see, why one cannot determine, which photon impacts the mirror first. It’s right, that the time in fast moving object runs slower. But due to the fact, that both photons have the same speed, that is to say the speed of the light, the time for both photons elapses the same way. It doesn’t matter, how fast the mirrors are moving - relatively to each other.
So the result must be, that the photon, which directs to the away moving mirror will impact it later.
HOW it looks like for the observer is not relevant, because this is relatively and not absolutely.
January 13, 2004
If one considers only the outcomes occurring at one side of the setup in our experiment, these outcomes look completely random: 50% of the times the photon passes the beam-splitter (i.e. the half-silvered mirror), ad 50% not. It is impossible in principle to know how a single photon will “exactly” behave.
One can assume that an individual quantum event is very much like a choice: In fact, in our experiment the measuring devices are adequately referred to as “choice-devices”, because at these devices a choice between two possible values takes place. Suppose you are driving a car: you cannot know whether the car before you will turn to the left or to the right at the next crossing, because you cannot read the mind of the driver. I think that the individual quantum event can be considered an “individual act of creation” (as Anton Zeilinger says), just in the same sense as the driver’s choice between left and right is one: the decision “creates” a bit of information (obviously, in neither case there is creation of energy).
Then, an important conclusion of quantum experiments like ours is that the very idea of quantum randomness is connected with the idea of order and intention. Indeed, if one considers only the outcomes occurring at one side of the setup these outcomes look completely random. But if one compares the outcomes at one side with those at the other side, one discovers that there is intention, ordering, and dependence. Hence, intention and randomness are not incompatible at all, and can originate from the same cause.
It is true, however, that today we don’t know yet which are the rules allowing us to connect quantum randomness and intelligent actions on the level of the brain. But the very existence of nonlocality clearly suggests that such a connection is possible in principle, and to find how it works in our brain is surely one of the great scientific challenges for the coming years.
In our experience, a laser beam illuminates a so-called nonlinear crystal, in which each photon of the laser converts down to a pair of photons. One photon of this pair is guided to a measuring device at the right side of the setup, and the other photon is guided to a measuring device at the left side. Both measuring devices (and the corresponding beam-splitters) are 55 m separated from each other. Each of them has a switch the physicist can arbitrarily set in different positions (a, b, c,…) to regulate the functioning of the device. One can arrange the things so that if both switches (the left side switch and the right side one) are set on the same position, one has the situation in which both photons behave exactly the same way and, therefore, the outcomes are only concordances (i.e. both photons pass or both doesn’t pass). The explanation you propose in your question is the same Einstein proposed: The photons behave like twins that carry a same genetic program and show the same behavior when they meet the same environmental conditions.
Suppose now a physicist on the left side switches on position a, and a physicist on the right side on b. Then one gets also some discordances (one photon pass and the other doesn’t pass). The physicist John Bell proved in 1965 that if Einstein’s hypothesis is true then the statistic of the outcomes should fulfill a mathematical relation now called Bell’s inequality: The probability to get a discordance with the switches in positions (a, c) plus probability to get a discordance with the switches in positions (b, a) cannot be less than the probability to get a discordance with the switches in positions (b, c). A number of experiments performed in the last 20 years show that nature violates this inequality. This violation proves that the production of the photons is NOT the key to the same behavior. We have to admit some kind of direct link between the two particles and, therefore, Einstein’s hypothesis is wrong. By the way, one can also arrange experiments using photons that come from different sources.
You mention very appropriately another possibility to escape this conclusion: the universe could be pulsing, so that all half-silvered mirrors in the universe are already connected even before the photons leave the source. But this possibility has also been excluded in the experiments: in fact, the setting of each switch occurs when the photons are already in flight, immediately before they arrive to the measuring devices, so that only links faster than light could ensure a connection between the beam-splitters.
One can attribute a time frame only to objects with rest mass. Photons have no rest mass and have no clock. Therefore it does not make sense to say: “the time for both photons elapses the same way”. To establish a time ordering you must take a clock, but to define a clock you have to use the state of movement of some massive object. Since in our experiment the choices are supposed to occur at the half-silvered mirrors, we determine the chronology of the events by means of the clocks associate to such mirrors. Through adequate distances and velocities we arrange the “before-before” situation, in which according to the clock of the half-silvered mirror at the left side the measurement at this mirror takes place before the measurement at the right side, and according to the clock of the half-silvered mirror at the right side the measurement at this mirror takes place before the measurement at the left side.
Since even in the “before-before” situation the correlations don’t disappear, we have to admit, as you suggest, that someone is calling: “Is there another photon anywhere in the universe which will pass a mirror?”, and if it is, he creates a dependence between the two “passing events” that astonishingly doesn’t correspond to any time ordering.
Center for Quantum Philosophy
Top of message
Top of page
March 17, 2003
I read an article in the French magazine "Science et Vie" about your experiment with entangled photons. It was very interesting.
I'm not a professional physicist; I'm a computer professional. What I thought when reading the article is very simple: it made me think on coupling. Imagine 2 parallel electrical circuits A and B with switches to light either a green bulb, or a red bulb. This means that A can have 2 colors Green or Red, and B also can have 2 colors Green and Red. If both switches are coupled together, A and B will be Green at the same time, or Red at the same time.
Now, in the computer business, we are used to think in layers. For instance, for networks, there is the ISO model with 8 layers. The lowest level is hardware, and the highest is application. This can be the same in physics. The hardware level can be seen as 2 separate devices, but 1 layer above, it can be seen as only 1 channel. You couple the switches, and then, you cause A and B to light either the two green or the two red bulbs. Internet is almost based on this. You don't know exactly what is going on at the hardware level, but the information you require at last reaches your display. It goes trough switches, routes are operated through whatever links, but you don't see it, and if the response time is quite fast, you don't care.
In your experiment, I feel that both light circuits are coupled, but what is done at the upper level is the result of quantum behavior. In fact, at the hardware level, both photon react and belong to the same "information", this one being at an upper level.
March 27, 2003
Your comparison with two coupled switches that cause to light either two red or two green lamps can be understood in two different ways:
a) You may mean that the devices A and B are the hardware level, and there is some program governing the switches and determining the behavior of the lamps. You could also put the example of two television apparatuses in two distant places switched on the same TV channel: They show the same image because they receive the same signals coming from the same emitter; the images on the two apparatuses have a common cause in the past. This is the way Einstein interpreted things. This explanation has been ruled out by a number of experiments since 1982, the so-called Bell experiments. These experiments test a mathematical entity called Bell inequality (according to the discoverer John Bell). If the correlation of the outcomes were the product of some hidden program, the results should fulfill such inequality. However experiment shows that the correlations violate this inequality. Therefore, one has to give up Einstein's explanation.
b) You may alternatively mean that there is no common cause in the past, but "one being" at an upper level connecting DIRECTLY the devices A and B, even if these appear to be separated at the lower level. But our experiment shows that the dependence between A and B the correlations reveal does not correspond to any real time ordering, and cannot be described in terms of "before" and "after". Therefore, the "one being" at the upper level seems to act like a powerful ordering mind capable of relating A and B without the flow of time.
Center for Quantum Philosophy
Top of message
Top of page
March 5, 2003
In cor/030302a you write: The quantum measurement means a choice between two possible outcomes, say between the values - and +. A choice has no duration and, therefore, I think it is not correct to say that the measurement takes a certain time though a very short one (“bien que très bref, dure un certain temps”). In this sense measurement relates rather to irreversibility than to decoherence. Once you hear a click, it is impossible in principle to restore the initial quantum state. However, the click you hear is not the cause, but only the mark of irreversibility. At what precise moment the choice takes place, nobody knows to date, nobody can tell you today which are the conditions defining when the quantum invisible state collapses and produces classical visible information.
I may be wrong, but that's not what I understand from what I have read about decoherence. The only choice I see is the decision to measure such or such quantum property (polarization of a photon, spin of an electron, Rydberg energy level of an atom, etc...). Then one has to carefully build the measurement apparatus, to make the quantum object interact with it, and to look at the outcome (value - or + to simplify). This outcome results from the interaction between the quantum object and the measurement apparatus + its environment. This "decoherence" takes a certain time, and that's why I wonder how this decoherence of the first measured photon is "feeled" by the second one (roughly speaking, to be quite sure that one photon is measured first, it suffices to imagine that the first photon is measured after only 1 meter path, and the second one in another galaxy...).
My references are:
Jean-Michel Raimond, Atomes, cavités et "Chats de
Schrödinger", Revue du Palais de la Découverte,
Vol.25, No 249, juin 1997.
Roland Omnès, Comprendre la Mécanique Quantique,
EDP Sciences, 2000 (pages 179-213).
Wondering about your most interesting experiment, I think that maybe it should be analysed in the reference frame of the source. I know that there should not be preferred reference frames, but is that still true for this kind of experiments? And though I did not like very much Costa de Beauregard/Cramer's hypotheses about going backward in time, I am now tempted to interpret your experiment in the reference frame of the source and with their hypotheses.
March 19, 2003
Effectively there is the choice the physicist makes “to measure such or such quantum property”. Quite important in nonlocality experiments are also the physicists’ choices of the settings of the measuring devices, as for instance the choice of the orientation of the polarizers or the phase of the interferometers: Bell’s theorem depends on the assumption that physicists are absolute free in selecting these settings. But the physicists’ choices are not “the only” choices involved in measurement. Nature itself also chooses. Considerer your gedanken-experiment with “the first photon being measured after only 1 meter path, and the second one in another galaxy”. We remember that each photon is monitored by two detectors, and denote D1(+) and D1(-) those monitoring the first measured photon 1: If D1(+) fires you get outcome +, if D1(-) fires you get outcome -. Since in each measurement the quantum object or wave always interacts with both detectors D1(+) and D1(-), the interaction alone cannot explain why it is that D1(+) fires and D1(-) does not (or the opposite). One has to accept that on the side of Nature a choice between D1(+) and D1(-) takes place. This choice of the outcome is irreversible, so that thereafter it is impossible in principle to restore the original quantum state. Suppose now you hear that the detector D1(+) clicks. Two views are possible: you can say that the irreversible choice of the outcome + (the “collapse of the wave function”) takes place because you hear the click, or say that you hear the click because the choice did take place before and produced a registered classical information. The first view leads to the well-known paradox of the Schrödinger cat. The second view, actually Bohr’s view, means that the proper measuring instrument cannot be treated as part of the quantum object under investigation (there is no Bohr’s cat!). It seems to me that decoherence refers to the quantum process occurring just before Nature makes its irreversible choice of the outcome. Anyway, assumed one accepts (as I do) the second view (i.e. Bohr’s position), nobody knows today which are the conditions that prompt Nature to make a choice (even if you know that such a choice has been made because you hear one detector’s click). As I wrote in cor/030302a, quantum measurement can be compared to some extent with the establishment of death: The irreversible breakdown of brainstem, does not certainly occur because a doctor establishes it on the basis of a number of observations; it occurs certainly before, though nobody can tell exactly when. But once the breakdown occurs it is impossible for us to bring back the dead to the previous living state. Suppose you adopt the view that the fact that D1(+) fires and D1(-) does not is the result of a choice. Then what photon 2 “feels” is precisely this choice of photon 1’s outcome, and not the decoherence the corresponding quantum state undergoes. It would be most interesting to hear what Jean-Michel Raimond and Roland Omnes think about this point.
Regarding the interpretation of our experiment by means of “backwards causation” (Costa de Beauregard/Cramer's hypotheses) you refer to, it seems to me that it leads to a big problem in situations in which the two measurements lie time-like separated, as it is the case in the situation you describe (“the first photon being measured after only 1 meter path, and the second one in another galaxy”). In this case “backwards causation” means that the measurement of photon 2 determines the outcome of photon 1. But this would imply that at the time T, at which photon 1 is measured the outcome of this event 1 is determined by another event 2 (the measurement of photon 2), to which at time T one cannot attribute any existence at all in any possible inertial frame (of the source or other), for the event 2 lies in the future light cone of the event 1. I think that this does not make sense. Actually, “backwards causation” is the strongest form of the bias that causality is always bond to the flow of time (you could also say the bias that all causal links have to be essentially Lorentz invariant): If you cannot bind causality to the time flowing forwards, you bind it to the time flowing backwards! As far as such a view does not imply observable consequences there is no problem in maintaining it. Nevertheless according to Olivier Costa de Beauregard backwards causation should have observable implications in experiments involving subjects who enjoy psi-capacities. In experiments with entangled particles, such psi-subjects are supposed to be capable of influencing on purpose the probability to get a particular outcome, say +, on one side of the setup, and because entanglement they would also change faster than light the corresponding probability on the other side. This would imply the possibility of signaling faster than light and, therefore, of changing the past. As I have been told such experiments are in preparation, so let us await the results.
Center for Quantum Philosophy
Top of message
Top of page
February 09, 2003
Here is a question about the relation between decoherence and non-locality.
It is now generally acknowledged, after Haroche's and Raimond's experiments, that the quantum measurement problem (wave-packet reduction) is well explained by the theory of decoherence.
However, decoherence is not an instantaneous process. In the case of measurement of the polarization of a photon, like in Aspect's or Gisin's experiments, what happens to the second photon when the first one is measured? According to the theory of decoherence, the first photon interacts during a certain time (very short, but that does not matter) with its environment. How is this interaction transmitted to the second photon, and how can this second photon transmit this "information" to its own environment?
I discussed that with a couple of French physicists (for example Herve Zwirn) but there was no conclusion.
(co-author, with Sven Ortoli, of "Le Cantique des quantiques") February 10, 2003
You have surely noticed that, in my previous mail, I keep the "conventional" interpretation of Aspect-like experiments, i.e. that one of the photons is measured "first", or, if one prefers, that there is a "before" and an "after". I think that the problem of the relation between decoherence and non-locality is better understood with this conventional interpretation.<
This problem was already pointed out in the postscript to the last edition of "Le Cantique des quantiques" (La Decouverte, 1998): (in French, but I think that you speak French)
"Le mystère de la mesure quantique est sans doute éclairci, mais cela ne fait que renforcer celui de la non-localité. En effet, la décohérence, quand on en fait la théorie, n'est pas un phénomène instantané. Et dans une expérience du type de celle d'Aspect, pour certaines orientations des détecteurs, c'est le premier photon détecté qui fixe la polarisation de l'autre, qui peut en être très éloigné. Le processus de mesure qui, bien que très bref, dure un certain temps, a sa contre-partie ailleurs dans l'espace".
Jean-Pierre Pharabod cor/030302a
March 1, 2003 You address the interesting question of the relationship between two main mysteries of quantum mechanics: measurement and nonlocality. I would like to begin with a comment on measurement. The quantum measurement means a choice between two possible outcomes, say between the values - and +. A choice has no duration and, therefore, I think it is not correct to say that the measurement takes a certain time though a very short one (“bien que très bref, dure un certain temps”). In this sense measurement relates rather to irreversibility than to decoherence. Once you hear a click, it is impossible in principle to restore the initial quantum state. However, the click you hear is not the cause, but only the mark of irreversibility. At what precise moment the choice takes place, nobody knows to date, nobody can tell you today which are the conditions defining when the quantum invisible state collapses and produces classical visible information. To a certain extent you can compare this state of things with the situation around the definition of death. Death, doctors say, means the irreversible breakdown of the brainstem functions. To establish death physicians have a checklist, the most important item of which is the lack of spontaneous movements, first of all spontaneous breathing. Death, the irreversible breakdown of brainstem, does not certainly occur because a doctor establishes it, but nobody can tell when it does exactly occur. Up to now nobody knows how much harm exactly defines the irreversible breakdown of brainstem. In any case, the conviction that death is irreversible means that we accept our incapability of mastering, even in principle, certain natural processes. Similarly quantum physicists admit that in measurement something happens that goes beyond our capability of restoring. The conditions required in order that a measurement takes place are not yet clearly defined: we don’t know whether measurement is determined by a still unknown new constant of nature, or by some mathematical impossibility theorem. I come now to nonlocality. Effectively, as you say, the "conventional" interpretation of Aspect-like experiments is that one of the photons is measured "first", or, if one prefers, that there is a "before" and an "after". John Bell used this explanation, and doing so he came to discover the astonishing quantum-mechanical nonlocality. This way of explaining things fits perfectly well for all experiments in which all apparatuses are standing still in the laboratory frame. Since the emission time of the photons is not exactly the same, and the fibers guiding the photons from the source to the measuring devices don’t have exactly the same length, according to the clock defined by the laboratory’s inertial frame, one of the measurements always takes place before the other, and the particle arriving later can be considered to take account of the outcome of the one arriving before. How is the interaction of the first photon with its environment transmitted to the second, you ask. If you assume that measurement means choice, then no interaction is transmitted but only information: When the first photon is measured an outcome is selected at random, and when the second photon is measured the outcome is selected taking account of the choice done before for the first photon. As you know the quantum mechanical formalism forbids us to use the link between the two measurements to phone faster than light. This means that this link is an unobservable one.
I tried to extend the “conventional” view to experiments with apparatuses in motion, in which two different clocks watch the arrival of the photons. The experiments prove that such an extension fails: the dependence the correlations reveal has no temporal counterpart. We have obliged Quantum Mechanics to tell us what its formalism means. In this sense the mystery of nonlocality has been clarified to some extent. By contrast, it seems that we are still far away of understanding why and when the irreversible choice in a measurement process takes place.
Center for Quantum Philosophy
Top of message
Top of page
February 07, 2003
In http://www.cfqp.org you write:
”Recent experiments have put in evidence that the correlations caused by a two-particles quantum entanglement cannot be described in terms of "before" and "after": The time-notion makes sense only in the domain of the relativistic local phenomena.”
This is certainly false, as proven by Bohmia폴ļmeans.br>
As far as I understand, you do not question the agreement between the predictions of BM and standard QM as well as the agreement of the predictions of QM with observation. As well, in BM quantum entanglement is described in terms of "before" and "after".
You may not like BM, many do. But "you don't like the description X" certainly differs from "cannot be described".
March 1, 2003
Many thanks for your comment.
First of all let me say that I like Bohmian Mechanics (BM), as John Bell also did. In fact, Multisimultaneity is nothing other that applying consequently the spirit of Bohmian Mechanics to experiments with apparatuses in motion.
Bohm's theory has been the first attempt to cast nonlocality into a temporal scheme. It uses a unique preferred frame or absolute time, in which one event is caused by some earlier event by means of instantaneous action at a distance. As you state Bohmian Mechanics makes the same predictions, as Quantum Mechanics. Nevertheless, if one tries to cast nonlocal causality into only one preferred frame, then it is not more reasonable to connect a "cause" event to an "effect" event in that frame rather than in some other frame. Effectively a single preferred frame ("quantum ether'') is "experimentally indistinguishable", John Bell said: The predictions would remain the same even if one assumes that the preferred frame is a virtual entity changing from experiment to experiment.
For this reason one is tempted to think that Bohm introduces absolute time just because he wishes to justify a causal description, but in the end, an untraceable "quantum ether'' is essentially the same as deciding arbitrarily which event depends on which one. What is more, in the particular case, possible in principle, of both measurements taking place at exactly the same time in the preferred frame, the only way of establishing which event depends on which is by arbitrary decision. Actually, Bohmian Mechanics (and Quantum Mechanics too) can be considered a causal description but not a real temporal one, and to date it has not lead to experiments capable of testing the timing-independence of quantum entanglement.
Taking account of your comment, I would like to precise the meaning of what I wrote as follows: Recent experiments with relativistic setups prove that quantum entanglement cannot be described in terms of “before” and “after” by any set of real clocks or experimentally distinguishable inertial frames.
Center for Quantum Philosophy
Top of message
Top of page