I've just finished the history book "The Theory that would not die", as mentioned here a couple of weeks ago [1]. It's got some interesting stuff to say on how Shannon's work emerged from the cryptography scene. According to the book, Information Theory was the culmination of all sorts of pragmatic techniques that were developed during code breaking. The terminology and expression was different (eg. information was measured in bans rather than bits), but it was really interesting to read about how practice developed into theory.

[1] https://news.ycombinator.com/item?id=11983200