In geology, the slab gap hypothesis is one of the explanations put forward to explain several instances of extension that have seemingly paradoxically occurred near subduction zones (which otherwise would typically compact crust, not stretch it out).
The hypothesis is that a gap forms between the two now subducting chucks of oceanic lithosphere when the spreading zone (the divergent plate boundary) of an oceanic plate is subducted under continental lithosphere. Plate tectonic theory states that divergent plate boundaries are fed by mantle upwellings. The slab gap hypothesis suggests that the mantle upwelling, unable to cause a spreading zone, causes the overlying continental lithosphere (and the continental crust on top of it) to spread apart. Note these mantle upwellings are much larger than any putative mantle plume responsible for making hot spots.
The slab gap hypothesis goes on to state that the upwelling can form very deep cracks, which in turn lets very fluid basalt lava to quickly spread over the land surface forming shield volcanoes and vast volcanic plains. If, however, the extention is spread over a very large area then these flood basalt events may not occur.
This idea has been used to explain the extension and very large flood basalts that occurred in what is now southern Washington, Oregon and northern California about 17 million years ago (see Columbia River Plateau). Slab gap has also been used to help explain the earlier creation of the Basin and Range Province.
Slab Gap web page at University of Colorado