Does prove always have a little bounce on the end? Or does this only occur when it is used as a noun (proof); that is, the bounce is part of a nominalizing "reduplication" (as in SIT versus CHAIR)? -Eric Gumpert