John Bell
Publications Physics
Publications Anthropology
Contact us

The Story behind the Experiments

John Bell

During 1988, I came in contact with John Bell in the CERN, the European Laboratory for Particle Physics in Geneva. With his support and collaboration, I launched the activity of the Center for Quantum Philosophy, promoting talks in the CERN and in various European universities. These activities aimed to stimulate the discussion on the philosophical and scientific challenges raised by the quantum theory, and show that this theory is now an integral part of knowledge and culture. In the CERN held talks among others: Herwig Schopper, John Bell, Constantin Piron, Nicolas Gisin, Asher Peres, John Ellis. Particular mention deserves a colloquium in Cologne (Germany) on May 13-14, 1990, organized in collaboration with the Lindenthal-Institute and the Institute for Theoretical Physics of the University of Cologne, at which participated two major exponents of Nonlocality research: John Bell and Anton Zeilinger (see [1] in German).


During a colloquium organized in the CERN on January 22, 1990, John Bell has been asked whether he thought that relativity and quantum mechanics could be incompatible. John Bell answered:

Photo courtesy GAP UK  http://www-gap.dcs.st-and.ac.uk"No, I can't say that, because I think someone will find one day a way of saying that they are compatible. But I haven't seen it yet. To me, it's very hard to put them together, but I think somebody will put them together, and we'll just see that my imagination was too limited." (see [2]).

This answer encouraged me to investigate in depth the tension between quantum mechanics and relativity. Quantum mechanics predicts correlated outcomes in space-like separated regions for experiments using two-particle entangled states. Bell-type experiments, conducted over the past two decades, suggest a violation of local causality: Statistical correlations are found in space-like separated detections: two photons, far away from each other and obliged to choose between two values (say between + or —), choose the same value, and we get either the joint outcome [+,+], or [—,—]. Violation of Bell’s inequalities ensures that these correlations are not predetermined by local hidden variables. Entangled particles behave as if there was a faster than light connection between them (see experiments).



I was strongly convinced that it should be possible to give a time-ordered causal explanation of nonlocal correlations, in terms of "before" and "after". Indeed, such a description is very well possible in conventional Bell experiments, in which all apparatuses are standing still in a 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. In fact, this is the way Bell tried to explain things, and, in doing so, he came to discover quantum nonlocality (see [2] and [1]). Orderings with one measurement before and the other after in time are referred to as before-after or after-before timings. In experiments with all measuring devices at rest, it is possible to explain quantum correlations through time-ordered causality.


But what about experiments with moving apparatuses in which several relevant frames are involved? In this case, different clocks watch the arrival times, and what is “after” according to the laboratory clock may become “before” according to one moving clock. Then, it is possible to define other time orderings: If each measuring device in its own reference frame is the first to select the output of photons, we have before-before timing. If each measuring device in its own reference frame selects the photon output after the other, we have after-after timing. Is it also possible to give a time-ordered causal explanation for relativistic experiments using apparatuses in motion? I assumed it was, and developed an alternative nonlocal description termed “multisimultaneity”, in which the time-ordered description of the nonlocal correlations extends to experiments with before-before and after-after timings. Consider, for instance, experiments in which the measuring devices are in motion in such a way that each of them, in its own reference frame, is the first to select the output of photons (before-before timing). Then each particle’s choice will become independent of the other’s and, according to multisimultaneity, the nonlocal correlations should disappear. In contrast, quantum mechanics requires that the particles stay non local, correlated independent of any timing, even in such a before-before situation.


As a matter of fact, the available conventional experiments at that date did not allow to decide between multisimultaneity and quantum mechanics. A new experiment using apparatuses in motion was needed.


 Monique et Marcel Odier

Vignette OdierSince it appeared rather difficult to obtain public funding for such an experiment, in 1992 I did a fund rising and was introduced to Marcel Odier, ancient private banker in Geneva, a man with a profound interest for the metaphysical implications of quantum mechanics. Since intuitively he shared the view that the concept of time does not apply to the quantum phenomena, he became immediately interested in the project and, together with his wife Monique, ensured sponsoring within the frame of the Fondation Odier de Psycho-physique. As scientific counselor of the Foundation Olivier Costa de Beauregard did acknowledge the relevance of the experiment and accompanied the project all along with great engagement and stimulating comments. The picture on the left shows Monique and Marcel Odier, Anton Zeilinger and Mrs Nicole Costa de Beauregard during a dinner in Innsbruck (click to enlarge).


Philippe Eberhard and Juleon Schins

On June 9, 1994, I exposed my ideas for the first time in a seminar in the CERN with the title: "Do the EPR correlations depend on the choices of the inertial frames?". Philippe Eberhard received in Berkeley the talk's abstract published in the CERN-bulletin no 23/94, and immediately contacted me. His theory, though for other reasons as Multisimultaneity, also predicted a disappearance of the correlations under certain conditions, and he proposed to study whether a common experiment could be possible. At that time I was in contact with Juleon Schins, who was working in Paris in the Laboratoire d'Optique Apliquée of  the ENSTA, and helping me very much in defining the terms of a possible experiment. Philippe, Juleon and myself met together at Geneva invited by Marcel Odier on April 2, 1995. 


Anton Zeilinger

After our Geneva meeting, both Philippe and Juleon had occasion at different times to visit the laboratory of Anton Zeilinger  in Innsbruck, and both discussed with him the possibility of carrying out the before-before experiment. Anton became interested and invited me to present my proposal to his group. During 1995 and 1996 I had very intensive and enjoyable discussions with him, Paul Kwiat and Harald Weinfurter, who at that time were working in Anton's group. On December 5-6, 1995 Anton cordially received Monique and Marcel Odier, Olivier Costa de Beauregard, and myself  to speak about possible financial support.  In Anton Zeilinger's group I got very good advices regarding the implementation of the required experimental environment. However it was more and more clear that the relativistic timing before-before required distances between the two measuring apparatuses substantially larger than those used in Innsbruck.


Valerio Scarani

In order to submit the experiment to open criticism of the scientific community I prepared a paper in collaboration with Valerio Scarani, presenting an experiment with devices in motion, capable of testing quantum mechanics against multisimultaneity. The article became published in July 1997 [Phys. Lett. A, 232 (1997) 9-14] (see experiments). We reached the conclusion that the before-before experiment could be realized with reasonable velocities (about 300 km/h) provided the measuring devices were separated from each other by a distance of about 10 km. So I visited John Rarity and Paul Tapster in Malvern (UK) in September 1997. They had done a Bell-experiment using a 4 km optical fiber and, though the real separation between the apparatuses was only of some meters, I thought I could convince them to put the apparatuses separated by several km.


Nicolas Gisin

Just before leaving to UK in August 1997 I got the new that Nicolas Gisin (GAP, University of Geneva) and his group had performed an experiment which demonstrated quantum correlations over 10 km, a distance ensuring the conditions required to realize the experiments that I proposed. In Malvern John Rarity confirmed that the collaboration with Nicolas Gisin would be the most convenient solution. Back from UK, on October 7, 1997 I met Hugo Zbinden and Wolfgang Tittel of GAP in Geneva for a first discussion. They became strong motivated and transmitted their enthusiasm to Nicolas Gisin. So on December 12, 1997, Marcel Odier, Nicolas Gisin and myself signed a convention in order to perform experiments using relativistic timings, and especially the so-called "before-before" timing. Then, Nicolas Gisin, Hugo Zbinden, Wolfgang Tittel and André Stefanov begun to work. I personally actively collaborated in the realization. The picture on the left shows the (visible) team behind the experiments (click to enlarge).

First of all we had to decide which components of the measuring devices should be set in motion in a real experiment. Since each outcome implies a choice between two values (+ or ), one must set in motion the device that is supposed to be that in which the choice of the outcome value takes place (choice-device). This means that if one invokes a description assuming quantum collapse at detection, one has to put the detectors in motion; if one invokes a description without quantum collapse (in the line of Bohm’s theory), the monitored beam splitters. Since to put beam-splitters in motion involved great technical difficulties, Nicolas Gisin and Hugo Zbinden decided to begin by putting detectors in motion. 

On November 26-27, 1998 we organized in Geneva a colloquium with Ian Percival, Sandu Popescu, and John Rarity to evaluate the work in progress. All of them agreed that the experiment with moving apparatuses was crucial to investigate the tension between Relativity and Quantum Mechanics.

In February 2000 we got the first results of the experiment using one detector in motion. They were in agreement with quantum mechanics (see experiments).

Hugo Zbinden - André Stefanov

Meanwhile Hugo Zbinden had an excellent idea: to take advantage of the fact that traveling acoustic waves can act as moving beam splitters, and use them to realize “before-before” experiments. After a technological feat, André Stefanov did them in June 2001. On Friday, the 22nd of June 2001 in the morning, I was in Geneva for the Colloquium where André Stefanov presented the first results he was obtaining. These refuted multisimultaneity, and I got the impression I was assisting to my burial. After lunch André and myself went to the laboratory to control the setup, and we found that, because of an error in the orientation of the measuring devices, the experiments had been done with “before-after” instead of "before-before" timing. So, in the annual meeting of the Foundation Odier in the evening of that Friday, no conclusive results could be presented, and the hope to beat quantum mechanics remain fully alive. On Monday the 25th, André restarted the measurements with the correct configuration. On Tuesday the 26th at 19:15, I suddenly became aware that my confidence in beating quantum mechanics was the product of a prejudice: I was assuming that causality always sticks to time. But nothing speaks against the idea of phenomena being produced by causes that are not bound to the limits of space and time. I then understood that this is the kind of causality behind the formalism of quantum mechanics. Next morning I called André, and communicate him my new expectation, that the correct measurements would confirm quantum mechanics, as it actually happened.


The final results

The final results of the experiments with moving measuring devices (see experiments) rule out the possibility to describe the quantum correlations by means of real clocks, in terms of "before" and "after"; nonlocal quantum phenomena cannot be described with the notions of space and time. This means that there is no time ordering behind nonlocal correlations, so the causal order cannot be reduced to the temporal one. Quantum correlations somehow reveal dependence between the events, or logical order. Experiment shows that this dependence, or logical order, is beyond any real time ordering. In the realm of the nonlocal quantum phenomena, things come to pass but the time doesn't seem to pass here.

This wonderful adventure had not been possible without the courage of Nicolas Gisin and Hugo Zbinden for undertaking non-mainstream research, the experimental capabilities of André Stefanov and the generosity of Monique and Marcel Odier.

Antoine Suarez
2 April 2003

Top of page

About the author

Dr Antoine Suarez took his Ph.D. in natural science at the Swiss Federal Institute of Technology (Eidgenössische Technische Hochschule Zürich or ETH) in 1975. While at ETH, he not only became interested in the philosophical significance of quantum mechanics but also in genetic epistemology. For more than a decade, he was engaged in research on cognitive growth that led to the development of improved methods for teaching mathematics and science to children. He directed the Swiss think tank, Institute for Interdisciplinary Studies (IIS), from 1985 to 1993, and, with major support from the Leman Foundation, he undertook studies that brought the insights of philosophers, theologians, and ethicists to bear on advances in science. Antoine Suarez was the first (with Valerio Scarani in 1997) in proposing experiments using moving measuring devices to investigate the tension between quantum mechanics and relativity, especially whether there is a real-time ordering behind nonlocal influences. He actively collaborated with Nicolas Gisin’s group in the realization of these experiments. In addition to articles in scientific journals, chapters in volumes of collected works, and an early study on the relation of thought to action in adolescents, Dr Antoine Suarez is the co-editor, with Alfred Driessen, of the book Mathematical Undecidability, Quantum Nonlocality and the Question of the Existence of God (Kluwer, 1997).

  Top of page


For any query or feedback on this site, feel free to send mail to the Webmaster.

Last updated on 26 February 2015.