- The event horizon is a two-dimensional sphere and, being two-dimensional, has zero mass and cannot exert any gravitational force.

- The black hole within the horizon is a three-dimensional massy object and can and does exert a lot of gravitational force.

- Assume at equilibrium our black hole is somehow exerting gravitational forces on its surroundings which are what you would predict if you accurately knew the black hole's actual mass.

- Assume the black hole moves, e.g. because of inertia.

- Now it should be exerting more force than previously on one half of the universe (the half it moved toward), and less force on the other half.

- Say it moved toward you. After a speed-of-light delay, you should actually perceive more force on yourself towards the black hole. But this can't be because a messenger particle was transmitted from the black hole to you. How can it be?

Assuming this shows that black holes cannot move seems unsatisfactory, given the recession of galaxies from one another, observations believed to show black holes colliding, etc. Where are my mistakes?

Followup: one black hole collides with another black hole of roughly ten times its size. Is it necessarily the case that the center of mass of the new, combined black hole ends up at the point that was the center of mass of the small-hole/big-hole system just as the small hole crossed the big hole's event horizon?

I find this concept of a black hole's surface having the contents inscribed on it to be really difficult for me to imagine; but I can take people's word for it, and it does explain away the paradox.

This reminds me that from some vantage points the universe seems so arbitrary sometimes. This is probably just my human intuition, but personally it just appears to me that nature is not really always elegant, but rather has these work-arounds and different layers to it to keep it working. Not unlike my terrible code.