Okay. Let's say that the hammer is moving rightside-up so that it will contact the bench handle-first. The hammer is really two distinct parts, the handle and the head. The head is usually significantly heavier than the handle. So we got this hammer moving, including the head, and then all-of-a-sudden it hits the bench. Naturally, the handle can't magically pass through the bench, so the handle stops on the spot. Now's the question: What happens to the hammer head? The head is not the handle--they're distinct--so it'll still be moving. Which direction will it be moving? Toward the handle. Now, the head is pretty heavy, so it takes quite a bit of effort to stop it. Additionally, all of its inertia (movement) goes into pushing the head onto the handle, so after the head stops, it'll be snugged up against the handle.
Now let's consider the other case. Let's say the hammer is moving upside-down and the head contacts the bench first. The bench stops the head (since the head also can't magically pass through the bench), but the handle is still moving. Now, the handle is significantly less massive than the handle, so it's easier to stop. Its inertia is also devoted into pushing the handle into the head, but since the handle is so much easier to stop, it doesn't snug up as much by the time it stops.
Basically, both cases snug the handle up to the head, but in one, the handle's inertia is used while in the other, the head's inertia is used. So it becomes a question of which has more inertia, the handle or the head?
Does that help any?
cookiemonster