Can Program Explanations Save the Causal Efficacy of Beliefs?

Frank Jackson and Philip Pettit offered the "program
explanation account� (PEA) in order to vindicate the causal
relevance of mental states such as beliefs. According to
J&P, a property F of a cause-event c (potentially a mental
property) can be causally relevant for an effect-event e"s
having property G because "e had G because c had F� is
an informative, non-redundant program explanation. If
PEA succeeded, the causal relevance of beliefs would be
vindicated and mental property epiphenomenalism would
be avoided.1 However, it doesn"t succeed.