Apparent paradoxes in Quantum Physics demand a sharp distinction between a
"real part of the world" and the realm of possibilities. In the former we m
ay distinguish individual elements (coarse events) relating to space-time,
with causal connections between them respecting the relativistic locality p
rinciple. The latter, quantitatively described by probability assignments c
onditioned on existing facts and depending on the definition of equivalence
classes of situations, cannot be subdivided in space-time categories. Ther
e are global correlations (Pauli principle...). The cut between the realms
of facts and possibilities implies an evolutionary picture of the "real wor
ld" in which the (generalized) anew of time assumes basic significance. Som
e deficiencies of existing theory are pointed out.