To summarize: correlation IMPLIES causation. It does not prove it. However, as I said, in natural science it is often all we have to go on, especially in biology. The mechanisms of biology are too complex and too numerous for us to understand them all. Even if our species survives for a billion years, I doubt we will connect all the dots in biology.
Not only that, but in biology it often happens that imperfect or partial correlation still implies causation. Ancient people knew that sex causes babies, even though it does not always cause them. Modern doctors long before 1964 knew that smoking causes lung cancer, even though it does not always, and even though the mechanism was not clear. Biology textbooks would be short if we only included fully-explained facts with well-established mechanisms. There are some beautiful examples of connecting the dots and establishing causality in biology. Especially Kotch's postulates, which I regard as the beginning of modern medicine. Many "double blind" modern medical studies almost seem to turn their back on causality and the kind of rigor Kotch introduced. They seem to demand only correlation. This is a pet peeve of Mike Melich, who sees the problem from a physicist's point of view. - Jed
