The simple interpretation of the San Andreas fault as a result of the collision between the Pacific and North American plates, after the consumption of the Farallon plate, conflicts with some known properties of the fault, especially in central California: The fault is too old, its displacement seems to be smaller than expected, and, most important, if it were simply the result of a collision of the two plates which it now separates, then Tertiary oceanic crust instead of Cretaceous continental crust should be west of the fault.
To reconcile these discrepancies it is proposed that the San Andreas fault originated before, and independently of, any direct interaction between the Pacific and North American plates. During its early history the fault delimited a sliver inland of an offshore trench. The sliver included the Salinian block and possibly also Baja California. When the two plates came first in direct contact, this was on a now virtually inactive fault zone offshore, along which strike slip occurred and triple junctions migrated. Inland, much smaller slip continued, or was resumed, on the favorably oriented San Andreas fault. Later, most of the movement was taken up on the San Andreas fault whose slip rate was greatly accelerated, until in the late Neogene or Quaternary it took up all the slip between the plates, and the sliver was attached to the oceanic plate.