Everytime Backus's Turing award lecture comes up I feel obliged to point to Dijkstra's take on it: (http://www.cs.utexas.edu/~EWD/transcriptions/EWD06xx/EWD692....) Reading both at once is just so much better.