The evolution of Earth models is reviewed and the open questions and proble
ms are highlighted. Generally, evolution of science was not linear, but pro
ceeded in "steps" of paradigms: where old ones remained within useful limit
s. "Geodynamic hypotheses", while embedded into the general concepts of spa
ce and time, were often mutually exclusive and competing until the 1900s. W
egener's concept of continental drift was the first successful globally uni
fying view. but it was discarded by most Earth scientists. The "real" chang
e of paradigms did not come before mid-century through geophysical observat
ions in paleomagnetism, seismology. bathymetry, seafloor geology and dating
, leading to the "New Global Tectonics"* of seafloor spreading and plate te
ctonics. Although real-time plate kinematics is now firmly established thro
ugh direct geodetic observations from radio-astronomy and with the aid of a
rtificial satellites., knowledge is incomplete and must be continually "upd
ated". Quite a number of problems are not yet solved. These may, some day,
lead to another change of paradigms, but certainly not back to "fixism". Pr
oblems are related to scale, frame of reference. and especially dynamics. H
owever, plate tectonics is presently the most successful theory in the hist
ory of geology. Further developments will be driven by discoveries of appar
ent dilemmas, contradictions and paradoxes. (C) 2001 Elsevier Science Ltd.
All rights reserved.