This sounds like the sort of thing advocated by Steve Mitchell in the pages of SIS but the story is at www.physorg.com/print221727791.html and derives from a paper that claims the crust can move up and down like a yo-yo. However, it is nothing like the sort of thing suggested by Steve Mitchell as they are referring to plate boundaries – although the idea might possibly be extended to whole sections of plate, assuming plates are not neccessarily huge affairs. The consensus view, or the one prevailing in recent years, is that subduction zones are areas when plates meet and merge, where volcanoes form, where lava is generated and where earthquakes occur. These are zones where the earth's crust is pulled down through convergence of plates, slowly descending into the mantlel like a huge conveyor belt, sinking into the interior and never to be emerge again. Australian researchers now say they have evidence, published in Nature Geoscience April 11th, that units of crust move independently, and much quicker than by hypothetical uniformitarian rates. They go up and down, and up and down again – and end up as part of mountain chains (a fact already recognised – in the Himilayas for example). The next question that might be aired is – if units of crust can move up and down at plate boundaries, as well as slide beneath each other, can earth's crust move up and down elsewhere?