Originally Posted by turbo-1

MOND is an ad-hoc correction of Newtonian dynamics that correctly predicted the behavior of low surface brightness galaxies at least a decade in advance of their observation. MOND may not have global applicability, but may instead be a marker of just how and to what extent Newtonian dynamics falls apart on galactic scales. This begs for epistemology. At what point could such an error have crept in? Are our observations in error? Are our assumptions about gravitational attraction in error? Could the fabric of space-time be polarized on large scales such that gravitational forces are not exactly the same in every domain? When you put all this on the table with the fact that clusters are far more gravitationally-bound than we expect given the calculated masses, it is hard to ignore that elephant in the room -- gravity on large scales does not act as we expect from Newton and GR. Since gravity is the dominant force on these scales, it seems rather cavalier to say that Newton got it right, and Einstein refined that, and then press on. Neither of them managed to get beyond a mathematical model of the effects of gravitation, though Einstein struggled for the rest of his life to lay out a mechanical model explaining how gravitation and inertia arise from matter's interaction with space-time. MOND has quantified the error of the Standard Model on galactic scales. If it serves no other purpose, it has served us well, indeed.