A scientific theory is a model of how the world works. A useful scientific theory is one that is able to provide predictions of what will happen in a given circumstance. (Look at the coverage of the avian flu for an example of how Darwin's theories have proven useful.)
When scientists say that a theory is "true," they are being sloppy. It would be more accurate to describe it as "useful." "Truth" belongs in the realm of philosophy or religion--science cannot tell us what is "true."
Some theories may well be accurate descriptions, but have not yet provided useful predictions. In the minds of several theoretical physicists, string theory is an example of such a beast. If string theorists come up with predictions that are matched by experiment or observation, then it will also be a "useful" theory.
Sometimes useful theories conflict in areas where they overlap. The most prominent (and intractable) current example is the contradiction between quantum theory and general relativity. A lot of very smart people have tried to identify a useful model that reconciles these two theories, and there are even a few interesting candidates for theories out there. These candidates are collectively known as GUTs (Grand Unified Theories).
Unlike theologians, scientist glory in these contradictions. When you find a really good contradiction, you are often on the cusp of learning something new and interesting.
--SCC
