Dear Sabine,
Nothing is basically wrong with reductionism and free will, provided we take to heart some provisions. I think all science begins with natural language (L), because he who does not know what distances, stones, rivers, weights, birds, stars, colors, seasons, etc. are cannot be a scientist. Scientists are people making claims (construct theories) beyond L, e.g. about H2O. In order not to remain castles in the air their theories must be grounded somewhere and the only place to ground them is in L. But L (to stay in the picture) is ignorant of atomic bonds, valences, molecular weights, joint electron orbits, etc. Great! Something has been ADDED to L Absolutely not contained in it - hence it cannot possibly be false. The problem begins only - as unfortunately is the rule - when scientists elevate H2O over L (known as epistemological inversion). What, however, shares no common measure can neither be represented by one of its parts nor can they be equated.
Et voilĂ ! The problem of free will vanishes, because it is no point physics has anything to say about.
Heinrich