Nuclear clustering has been a subject of intense study since the advent of
heavy ion accelerators. Looking back more than 30 years, we are able today
to see the connection between resonances in heavy ion reactions and extreme
states in nuclei. In particular, excited states close to the decay thresho
lds into substructures show pronounced cluster structures. The ever repeate
d question about the possibility to describe such cluster phenomena in an e
xtended (to a very large basis) and deformed shell model, has recently been
revived in view of improved computational facilities. The new experimental
facilities and the powerful new detector arrays for gamma-spectroscopy and
for charged particles have increased the sensitivity to study clustering p
henomena by orders of magnitude. Today we are able to study nuclear structu
re at the limits of the mean held concept. With the advent of radioactive b
eams and the study of nuclei, which are weakly bound in their ground states
, the question of the limits of mean held concepts has reached the low exci
tation energy region of nuclei.