🚀 go-pugleaf

In article <ksvk2iINNj20@mthvax.cs.miami.edu>, wb8foz@mthvax.cs.miami.edu (David Lesher) writes:
|> Recently I suggested that a book could substitute for a one-time pad.
|> I've gotten several pieces of mail claiming the method I was proposing
|> was trivial to break.  Others have said it is not.
|> 
|> So my query to the netexperts is: which is it?
|> 
|> The method I'd heard of:
|> 	Cleartext:	The Rain in....
|> 
|> 	1) Open book at pseudo-random. 
|> 	2) Close eyes, point to somewhere on page.
|> 	3) Open eyes, find nearest  'T'
|> 	4) Page 38, line 7, letter 37 
|> 	5) output 38737 or such as encrypted 
|> 	6) Mark page so you do not reuse it.
|> 	7) Loop to 1) for "h" 
|> 		and so forth.
|> 
|> Now, if you never reuse the page, or maybe the paragraph, have you
|> provided enough data so that someone without the same book can decide
|> it or not? I've gotten different answers.
|> 
|> {I can't remember where I heard of this, but it was in the popular
|> press somewhere, maybe a novel.}

I don't have my copy of Kahn's, The Codebreakers, with me, but as I recall, William Friedman broke a book cipher of this sort during the 1930s.  The cipher was in use by some kind of smugglers, and Friedman only found out what book they were using when they were finally brought 
to trial.   As I recall, the basis of breaking book ciphers is that
the book follows the statistics of the English language (or Spanish
or Urdu or whatever).   Can anyone on the net fill in the history
of this better than I can?

(Just for reference, William Friedman was in charge of the group that
broke the Japanese Purple code, prior to Pearl Harbor, and is generally
regarded as the father of modern cryptography.   His wife was also
an outstanding cryptographer, so she is presumably the mother of
modern cryptography!)