A Non-Factorable Local Realist Model for EPR Experiments with Timing Variations

The violation of Bell's inequalities in some EPR-type paired-photon experiments could be as much due to the sensitivity of the coincidence monitoring system to timing variations as to other causes of \enhancement". If there are systematic variations of processing time with polarisation angle, then the \factorability" assumption may not hold if window size and other parameters are not carefully chosen. The eeect is illustrated using the chaotic ball model introduced in an earlier paper. Some little known facts from Alain Aspect's thesis are presented. It is argued that, though Aspect took considerable care with his choice of parameters, he did not take into consideration the factors discussed here and the experiments may in consequence have suuered bias. Part of this story has been known for some time, but appears to have received little attention. Further experimentation is needed.