Strain is caused by differential motion of molecules on one side of the fault
vs. the other side of the fault. In other words, there is an actual distortion of the
material. Stress is simply a force which tends to (in this case) act to move something
like a fault.
Think of the San Andreas as an example. Deep in the earth(around 15-20 km) the
plastic, heated North American and Pacific Plates are sliding effortlessly past
each other. They are too hot to break, so like any plastic, they just deform. On the
surface, however, deformation is not possible without breakage. If the North America
plate is to move as a whole, the movement of the plastic layer below must at some
time generate a break (rupture) on or near the surface.
With that in mind, how can this be reversed? The easiest way is to reverse
the motion of the North American with respect to the Pacific Plate. Because the
plate is moving so slowly, there is very little inertia, so an instantaneous change
due to external forces is conceivable. I am referring to the possibility that a
large storm could change the motion of the North American plate wrt the Pacific
Plate and temporarily relieve strain between them. This could also be
accomplished by a meteor collision. Plate motion as evidenced by hot
spot tracks shows that plates make sudden changes in direction at the time
of meteor collisions.
In terms of far-field triggering, it is possible that areas which are at inter-nodes,
where seismic energy input is less could have strain reduced temporarily.
Finally, the most accepted method for doing so is in the general region of
other large earthquakes. Once a quake has happened, the differential strain
at the end of the rupture falls into two categories - half the areas will experience
an increase in strain, the others will experience a decrease in strain. Try this
mental exercise. Consider taking two books connected to two slabs of soft rubber at
the ends of the books. As you slowly slide one of the books past the other,
the rubber at the end of one is compressed and at the other end it is stretched.
Deformation is taking place in the process leading up to an earthquake. Keep
doing this until a good deal of deformation has occurred in the rubber slabs.
Now hold onto only one of the books and let the energy stored in the deformation
of the rubber be released (this is the earthquake). What has happened? The deformation
in some areas has disappeared in the earthquakes and strain throughout the
block of rubber has been lessened in some areas, decreased in others.
Not a perfect model, but play with it in your mind, and see what you can come
up with. Add a plastic layer beneath etc. Physics involves a lot of mind games,
so have fun thinking.

The one which is pushed into the pillow deforms (adds strain)