The established models of the structure and dynamics of a ''typical'' Hawaiian volcano lead inevitably to a paradox: finite element viscoelastic calculations, using TECTON <Melosh and Raefsky, 1980, 1981>, predict tht a compressional stress field should characterize their upper flanks; instead, extension is observed. The paradox is solved by postulating that volcanic evolution is determined by feedback processes related to growth and spreading of the volcanic edifice. Many examples of the process of spreading are documented in the geologic literature on volcanoes of a very broad size range and very different geologic settings. The process is similar across many orders of magnitude in size: the deformation of the substratum related to magma intrusion during lava dome building events is at the lower end of the scale; the spreading of seafloor at oceanic ridges is at the higher end of the scale. Volcanoes may evolve from small constructs into huge shield volcanoes and perhaps into ocean plates.

Key elements for volcanic spreading are (1) the existence of a weak basal layer and (2) a sufficiently high mass and magma influx to drive the process. As the mass of a volcano grows, it passes through five phases (which may overlap, repeat, or be omitted): building, compressing, thrusting, intruding, and spreading. During the building phase the mass of a volcano does not significantly affect the stress field of the substratum; primitive magmas are likely to be erupted. As the mass increases, the volcano subsides and becomes characterized by a compressional stress field which inhibits any further intrusion of magma into the volcanic edifice; magma may differentiate in the crust, allowing for large caldera-forming eruptions to occur. In the next phase, thrusting begins on a decollement usually located at the foot of the edifice; while the stress field on the upper slopes of the volcano remains compressional, the stress at the base, close to the center, becomes extensional, allowing the magma to intrude and differentiate at the base of the edifice; a basal intrusive complex begins to grow; explosive eruptions may be expected.

Finally, spreading of this complex creates an extensional stress field on the upper flanks of the volcano, confining compression to the lower slopes of the edifice; the extension allows primitive magma to erupt again at the summit. The process of spreading seems to have a fundamental influence on the structural and magmatic evolution of volcanoes. At every scale, this process results from coupling and feedback between the gravitational and thermal fields. Thus studying this process may improve our understanding of the relation between structural dynamics and magma evolution and of the origin and evolution of volcanic seismicity and eruptions. ¿ American Geophysical Union 1994