Quantum Correlations and Common Causes

It is investigated whether in principle it might be possible to give a causal explanation for the correlations that are observed in variants of the Einstein-Podolsky-Rosen (EPR) experiment. Bell's Theorem states that certain assumptions about the experimental setup and about causal explanations of the correlations in question imply the so-called Bell-inequality. This inequality, however, is experimentally violated. Therefore at least one of the assumptions from which it follows must be false. Contrary to the received view that the assumptions are very weak it is argued that at least the assumption about how to apply the Principle of the Common Cause to the EPR correlations is unreasonably strong. That this assumption - a so-called screening-off condition - is false would therefore not be surprising. However, it is proved that even if the screening-off condition is weakened the Bell inequality can still be derived. This derivation of a Bell-type inequality is distinguished from other derivations by the weakness of its assumptions, one of which must be false. Among the candidates for being false there is in particular the assumption dubbed 'no conspiracy', which asserts that certain causally independent event types are also statistically independent.