Wow, Narc, you really do have trouble understanding simple concepts. Here, let me repost something I just posted on another thread:
In a gravitational field, "up" is almost always* defined to point exactly opposite to the direction of acceleration due to gravity.
Mathematically, if u have a gravitational potential given by a function G(x,y,z) then "up" points in the direction of Grad G (usually written ∇G).
*The only exception would be in an accelerating system e.g. a spacecraft.
EDIT: Or you could define "up" differently if it would make things easier, like cbarnett said, but peple usually don't.
As you can see, on Earth "up" will always be towards the sky and away from the centre of the Earth. Objects will always fall "down". On the moon, "up" will be away from the centre of the moon.
Do you understand this definition? Have you got any problems with it?
Of course, this definition would only work in RET.