This is something I've always wanted to get off my chest as a former astrophysicist. Most readers can just tune out now.
USATODAY.com has an article reporting observations of the microwave background that purportedly support the inflation extension to Big Bang. I can't comment on the findings without reading the actual paper in ApJ, but the article includes the statement
Normal matter, the stuff of people and planets, is only about 4% of the combined matter and energy in the universe. Dark matter, invisible and exotic physical particles, and dark energy, a gravity-defying force behind the continuing expansion of the universe, makes up the rest.
This idea has always bothered me. Here's the story. Inflation cosmology is an attempt to solve some paradoxes in Big Bang. Its essential point is that in the first flicker of time after the Big Bang event, the universe expanded at speeds far greater than the speed of light. That explains the essential isotropy of the universe, how parts of the universe too far apart to have ever communicated can have similar structures and properties. That's known as the horizon problem. See Wikipedia's entry for
inflation for a more thorough introduction.
The detailed physics of inflation is far beyond me, but a critical piece of the theory is that the universe must be perfectly flat. This has to do with general relativity. Basically, if the universal density is high, the universe will collapse on itself. If is the density is low, the universe will expand forever. The transition point between the two is called a flat cosmology. (In GR terms, the metric tensor has constant elements only on the diagonal. Also known as the Minkowski metric.) The corresponding density is called the critical density. Inflation requires the universe to be at exactly this density. Unfortunately, observations give a density well under this critical value. That is, of course, only looking at observable matter. Since inflation requires the critical density, and observable matter gives a density which is only a small fraction of critical density, everything else must be totally unobservable matter, which we call dark matter. That's where USA Today's statement comes from.
This has always struck me as a modern example of epicycles. In the old geocentric views of the solar system, they had to impose bizarre, unphysical motions on top of the simple circular orbits the planets were assumed to have around the earth. These were called epicycles and were needed to explain things like retrograde motion. The only purpose for these epicycles was to bridge the gap between what their theory predicted and what was observed.
The dark matter hypothesis does exactly the same thing. It proposes that most of the matter of the universe is totally undetectable (meaning we can't actually test for it) whose only purpose is to flatten the universe to comply with the predictions of inflation. I call dark matter the cosmological fudge factor. Whatever we don't see, we just assume is there and unseeable. That's always struck me as totally unscientific.
The idea of dark matter comes from the study of galactic rotations. Rotation curves are measures of how fast matter is orbiting the galactic center as a function of distance. We can determine the mass by looking at brightness. Knowing the mass and the speed, it is fairly simply physics to predict what will happen. Unfortunately what was sometimes found was that there was not enough mass in the galaxy to keep matter in the outer parts of a galaxy in the orbits that were seen. Matter should have been flying off the galaxy rather than being bound gravitationally to the galaxy. The solution, again, is that there must be matter that we're not seeing that is increasing the mass, and therefore the gravitational pull.
Now, this use of dark matter is not so bad. First of all, unlike inflation, the theory we're trying to preserve is extremely well established (basic gravitation) so there's very good reason to expect its predictions to be right. Beyond that, it's simply not unreasonable that there would be matter in a galaxy we're not seeing. Dead stars, dim stars who are simply not bright enough to be detected in our telescopes, dust, etc. All are examples of mass, potentially considerable mass, that would not be emitting light for our telescopes to see.
But when this idea is applied to cosmology, the dark matter morphs from perfectly reasonable things to bizarre, exotic particles no one has ever seen, nor can they see by definition.
Like I said, I have always found this rather unscientific.