June 4, 1999
I suspect that many of the results we see are caused by the way the apparatus is made. It has been designed by people who were trained to believe quantum theory right, and they have, perhaps subconsciously, produced detectors, in particular, that simulate quantum theory. They give the effect that there are "photons" of light, all (for a given frequency) the same energy. If QT is actually true, the output from a detector should be proportional to the intensity of the input for low intensities, however low, as the number of photons is proportional to the intensity. A slight concession to reality could be allowed, in that it has to be recognised that the detectors will, in these low energy conditions and when set in "Geiger mode", always have a "dark count". They will produce the occasional output even when there is no applied input. Given an input, though, the additional number of outputs should be exactly proportional to its intensity. I am fairly sure that it is not. Experiments could hope to show this, though it is not so very easy as the only way to measure intensity is by using another detector!
An idea that I think would be fun is to try and extract information on detector responses by looking carefully at the results of EPR-type experiments. Fortunately, so far as discriminating between QT and realism is concerned, it is not critical to know what the detectors actually do, only what the combined effect of the polariser (or interferometer) and detector is - a much easier problem!
Until we (the public) know just what the detectors are doing, it is hard to judge what the other parts of the apparatus are getting up to. Quantum theory says that individual photons go one way or the other at beamsplitters, whereas classical theory, in its basic form, says they split in two halves, each with the original frequency but half the amplitude. Now a variant of classical theory - one that I think is likely to be right in this respect at least - is "Stochastic Electrodynamics" (SED), Trevor Marshall's speciality. This says that the two halved need not be equal. He argues that this is because other background radiation (what he terms the zero point field (ZPF)) is added to the signal at the beamsplitter. There is inteference, acting constructively for one output, destructively for the other.
I have an idea that the truth might be more a matter of a chaotic effect, with the beam oscillating between the two outputs as the material of the beamsplitter is coaxed into first one then another of two internal states. The whole matter could be investigated using variations on an experiment by Grangier, Roger and Aspect, 1986 (Europhysics Letters 1, pp 173-179), or I have my own variant.
In 1998 I suggested to Anton Zeilinger (head of quantum optics at Innsbruck) that he could try a simple-sounding experiment using two beamsplitters. He has not replied to my requests for a progress report, though it is now many months since he said he would put a student to the task.
"EPR" experiments obey local realism!
I've put together some of my ideas for refuting statements that "experiments confirm" that Nature does exhibit nonlocality and so forces us to rethink our whole philosophy. They don't. Nature operates locally. We can go back to the comfortable days of "naive realism" without any conflict with the evidence!
The experiments I propose should not only show that ordinary logic continues to reign supreme, but should help us to formulate our ideas on the nature of light, how it behaves in "atomic cascades", nonlinear crystals, lasers, optical fibres, photomultipliers ... It does not behave like particles.