Nobody can tell you WHY it happens, but there are two current theories of HOW it happens. Both are pretty good, they disagree with each other, and both have serious problems (although not if you talk to their believers!)
The oldest one is general relativity. It was announced in 1915 and has passed a lot of experimental tests since then. It explains gravity as due to curving geometry of spacetime. The "path of least effort" through this curvature is itself a curved line, such as an orbit. The problem with general relativity is that it is a classical theory, that is to say not a quantum theory. This is a constant bother to the physics community, but quantizing gravity has turned out to be a tough problem. We can't just leave it at that because even the relativitists agree that at the planck scale the classical theory fails. And many believe that the presence of mathematical singularities (black holes) in general relativity means it is an incomplete theory.
The newer theory is superstrings (which includes now branes and M-theory). PBS had the Elegant Universe on last week, I am sure they will rerun it, and that should give you an introduction to what supertstring theory is about. String physics explains gravity as due to the exchange of particles called gravitons. The string physicsists assert that their gravitons explain everything general relativity does, and the relativitists deny it.
The argument turns on whether "background independence" is an important feature or not. In general relativity, very crudely, the curvature is proportional to the local density of energy and momentum, including mass energy from e = mc^2. So spacetime is right in there, a part of the physics. But in string theory, like all quantum theories to date, spacetime is a background that isn't modified by the physics. So does that make a diffference? Time will tell.
The problem with superstring theory is that it only predicts things at too high an energy for our experiments and astronomical observations to detect. So it has no empirical support. GR, on the other hand seems to pass a new observational test every year or so. This is still going on; the LIGO experiment is set to test for gravity waves, a prediction of GR.
Many of the things I have said here should be modifies in light if ongoing physical research - see the strings branes and LQG forum where real participants in this research post. But this gives you at least a basic outline of the situatioon.
. but i appreciate all asnwers! And please: don't hesitate to call me an idiot for asking such a stupid question
