I've written about various evolutionary theories of music before. Now I point you to a good radio program from the BBC that interviews Steven Mithen, Professor of Early Prehistory at Reading University and Lawrence Parsons, Professor of Cognitive Neuroscience at Sheffield University, on this subject. Mithen, the author of The Singing Neanderthals, discusses the archeological record of music. I've read before about the 30,000 year old flutes, but Mithen makes an interesting claim. By looking at the anatomical design of vocal tracts in proto-humans, he feels that our ancestors were singing millions of years ago, before the development of language. Mithen bases this argument on the facts that there are no records of symbols used that far back, and that behavior was not languag driven. I'm not sure what he means by this, I'll have to read the book to find out. Parsons describes all the interesting research in functional imaging of the brain. Comparisons of the brains of musicians and nonmusicians, and scans of brains while listening to or performing music have revealed some fascinating data. Take a listen, the interviewer does a good job of keeping things interesting, understandable, yet rigorous.
(via Mind Hacks. Also note the article about hearing implants and musical appreciation.)