A straightforward comparison between the strain energy (in bending) input rate to the continuously developing outer rise system and the elastic strain energy release rate occurring in the form of normal-faulting earthquakes in the lithosphere offshore of oceanic trenches requires that extensional strains of ~ 10-2 that exist in the upper half of the 'elastic plate' indeed be elastic and thus that kilobars of deviatoric stress exist in the outer rise system. This result does not preclude anelastic strains up and beyond ~10-2 that can and almost certainly do accompany the outer rise deformation; it does, however, presuppose a steady, flexure origin of the outer rise, as is almost always assumed in modeling studies of bathymetry and gravity data. For the purpose to which it is put here, this assumption, at best, is likely to be valid only on a time scale corresponding to several cycles of strain accumulation and release associated with great underthrusting earthquakes onshore of the trench axis; the decade or so of reasonably well-determined and complete seismic observations that are presently available for the outer rise region represents only a small fraction of this time scale. 'Lithospheric normal-faulting earthquakes,' of which the principal examples are the Sanriku (March 3, 1933) and Java trench (August 19, 1977) earthquakes, cannot arise from tectonic stresses due to bending and thus certainly argue against the flexure origin of the outer rise locally and perhaps globally as well, despite the separate bathymetric, gravimetric, and seismologic evidence which provide substantial support for the flexure origin of the outer rise. At least in the case of the Sanriku earthquake it is unlikely that the entire lithosphere was fractured. In the case of the Java trench earthquake the supposition that the depth of faulting is no more than 30 km implies an exceptionally large stress drop.