Galaxies are strange beasts: simply put, they rotate faster than they’re supposed to. At the speeds stars orbit the center, their centrifugal force (okay, not actually a real force) should overwhelm gravity and send them flying off into space.
One way to explain this is to assume the existence of a kind of matter we can’t see, which significantly raises the total mass of the galaxy and therefore its gravitational pull. This, called the dark matter theory, explains some phenomena but struggles with others, while something that doesn’t emit, reflect or even absorb any kind of electromagnetic radiation just seems a little weird to some. Another hypothesis, which also works in some cases but not all, is called Modified Newtonian Dynamics, or MOND for short.
Newton Spinning in His Grave
It may seem absurd for scientists in search of knowledge to make up stuff apparently at random. Both of these somewhat kooky theories, however, fit the data we have reasonably well. The same can be said for the double-helix structure of DNA, the Schrödinger equation and how atoms arrange themselves into benzene rings. All of these involved a level of creative thinking that must have seemed insane, to some and at the time, but they could all be proven eventually, and each kicked off an entirely new branch of science.
It’s already well-known that size matters: subatomic particles behave completely differently to volleyballs. It’s also true that the laws of physics change when something is going very fast or under the influence of a very strong gravitational field. It is therefore possible that making a “small” alteration to the math we have to describe gravity, which only becomes relevant on very large scales, can eliminate the need for something like dark matter to exist.
How It Works
The basic idea behind MOND is that, when it comes to large distances and low accelerations, star motion can be explained better by either changing the famous F=ma to F=ma2, or writing the gravitation equation F=Gm1m2/r2 as F=Gm1m2/r instead. This has since lead to a number of more detailed interpretations of MOND, both relativistic and non-relativistic.
There’s no inherent reason to assume that this is accurate, but it does help to explain a number of things that otherwise require dark matter. It’s also worth pointing out that we’ve only really tested our current theory of gravity within the solar system; for all we know, there could be space gremlins further out who like messing with the laws of physics.
Arguments for and Against
In our solar system, Mercury’s orbit is the smallest and fastest at only 88 days, while Pluto didn’t make it around the sun once in the 76 years between being discovered and fired from its job as a planet. This makes sense: the more gravity acts on an object, the faster it accelerates.
With entire galaxies, however, the outer rim rotates at the same speed as stars closer to the center. Dark matter can explain this, but only by assuming that it’s spread around fairly evenly instead of crowding around the core like visible matter. This is what you’d expect something with mass to do. MOND provides a much simpler, more elegant and more accurate explanation…but only for individual galaxies. Once you get to a system the size of a galaxy cluster, it breaks down and dark matter seems to be the winner.
There’s also the matter of the cosmic microwave radiation, a leftover from the Big Bang. This is generally uniform, but varies slightly in intensity and color temperature. These tiny fluctuations, with the help of some pretty impressive software, can be analyzed to produce a kind of sky map. The results are consistent with something up there that’s the same density dark matter is supposed to be. MOND offers no comparable answer.
Overall, in terms of the ability to make (correct) predictions, dark matter seems to be the stronger theory. Most cosmologists (who are obviously most concerned with very large-scale systems) regard modified Newtonian dynamics as a kind of mathematical curiosity. Several people are still doing work on it, though. It could just be that distance really does affect gravity in more ways than decreasing it at an inverse-square rate.
Related Information:
Dark Matter
More about the standard (but still unproven) way of explaining why galaxies seem to weigh more than they should.
Astronomy
A little about how astronomical research is conducted.