In Kyudo, the bow is held in a very light, delicate manner called "tenoche." Given the spirit and intentions of Kyudo meditation, a forceful grabbing of the yumi is inconceivable. Now if one applies the delicate "tenoche" grip of Kyudo to shoot a symmetric long bow like in figure 2, this is what happens: The bow simply rips past the delicately placed fingers of the tenoche grip and drops to the floor. This has been observed by one of us (J.Neu) who is a three year (raw beginner) student of Kyudo.
The problem of handshock prompted us to ask, where does one place the grip of the bow so as to minimize or altogether eliminate handshock? The answer, from numerical experimentation, is, about 1/3 of the way from the bottom limb tip. When such an asymmetric bow is shot, the handle basically pivots in the hand, with almost no back and forth motion. The situation is analogous to the free vibrations of a beam. The center of the beam vibrates back and forth with a large amplitude, and there are two off-center nodes. The motion about each node is a mere pivoting, with no back and forth displacement of the node itself. Figure 3 depicts the time sequence of a yumi's configurations during the launching of the arrow, as determined by Tom's numerical simulation. The presence of the two nodes, and particularly the node at the grip, is clear.
http://math.berkeley.edu/publications/newsletter/1996/bowmakersdiary.html