"To Effectively Discuss Evolution, First Define 'Theory'" (R. Lewis, The Scientist, May 12, 1997, page 13) is pretty good advice. Even better, however, is to first define "evolution." The article uses the word "evolution" equivocally to stand for three distinctly different concepts: (1) change over time, (2) common descent, and (3) the Darwinian mechanism. These three ideas require radically different kinds of evidence to support them.
To conclusively prove change over time, one need only demonstrate that a type of organism (say, Tyrannosaurus) was alive in the past and is not alive today. Everyone, even the staunchest biblical fundamentalist, accepts change over time; pretending that this is the issue is to argue against a straw man. Unlike change over time, common descent cannot be conclusively proved. It becomes more plausible, though, as more fossil intermediates are discovered, showing one form blending insensibly into the next, as Darwin thought they...