Classical relativistic physics assumes that spatially separated events cannot influence one another (“locality”) and that values may be assigned to quantities independently of whether or not they are actually measured (“realism”). These assumptions have consequences—the Bell inequalities—that are sometimes in disagreement with experiment and with the predictions of quantum mechanics. It has been argued that, even if realism is not assumed, the violation of the Bell inequalities implies nonlocality—and hence that radical changes are necessary in the foundations of physics. We (...) show that this conclusion does not follow unless the locality hypothesis is strengthened in an implausible manner. (shrink)

It is shown that the Bell inequalities are conditions that must be satisfied by the probability functions of certain three-variable systems if they are to be expressible in terms of a single, nonnegative function. The generalized Bell inequalities, or CHSH inequalities, play a similar role for four-variable systems. The physical significance of the results is discussed.

The new arguments in the cited paper are shown to be as unsatisfactory as the old.