There is an interesting article on slate if peer review works. Read it here...