Collapse interpretations speculate that there is some mechanism which causes the wave function to collapse down to just one version. But the collapse process must satisfy certain conditions. If the wave function of a single photon collapsed too quickly, then the observed interference patterns in, say a double-slit experiment, would not occur (because one part of the wave function would have simply gone away and so it could not interfere with the other part). To avoid problems like this, it is speculated that collapse can occur only when the different versions contain many particles in different states.

Suppose for example we do a Stern-Gerlach experiment on a single silver atom with a detector on both paths. The detectors are assumed to have pointers on them, one of which points to 0 (no detection) at the end of the experiment and the other to 1 (detection). The pointers are presumed to have many billions of particles in them. So, the reasoning goes, the collapse would not occur until the pointers, with their many particles, register the result. And then it occurs very quickly, say in a millionth of a second.

Interference experiments. One way to look for collapse is to exploit the many-particles idea in interference experiments. One would do an interference experiment, similar to those done on photons or neutrons, for objects with more and more particles. If, for some number of particles, the observed interference pattern did not follow that derived from standard quantum physics—if it disappeared, for example—then this would imply that the wave function had collapsed to one branch or the other so there would be no interference of the two branches of the wave function. This has been done with Bucky balls (arrangements of carbon atoms) having about 720 protons and neutrons, and no deviation from standard quantum physics in the interference pattern was found. That is, the experiment provides no evidence for collapse.

There have also been Superconducting Quantum Interference Device (SQUID) experiments in which approximately a billion electrons took part. The interference between two states of the billion electrons was just what quantum physics predicted, so, again, there was no experimental support for collapse.

Unusual physical processes. There is a second type of experiment that can be used to look for collapse. If there is collapse, it can lead to energy fluctuations that occasionally (very occasionally) cause the ejection of an electron from an otherwise stable atom. In one experiment, a chunk of germanium was monitored for a year to see if there were decays at a particular energy. The negative results (no decays) are consistent with no collapse. But if it is assumed that collapse occurs, the null results imply that electrons and photons, in essence, do not contribute to the collapse process; only protons and neutrons contribute.

In another experiment that yields information on the possibility of collapse, the smallness of the disagreement between theory and observation in the Sudbury experiment on solar neutrinos is consistent with no collapse.

In summary,

[P16] Despite a number of ingenious and sensitive experiments, there is no evidence for collapse.