We show that from a study of collective motions built on highly excite
d states we can extract important information on open problems in nucl
ear structure and dynamics: i) Competition between one- and two-body d
issipation, ii) Effects of correlations and related density fluctuatio
ns, iii) Transition in the nature of nuclear collective motions: from
pure quantum to classical sound waves, iv) Pre-equilibrium effects: th
e ''dynamical'' dipole in charge asymmetric entrance channels, v) Fusi
on hindrance from dynamical instabilities. We try also to discuss the
relevance of new, more exclusive, experiments and of a larger systemat
ics.