"Another possibility which is still open for me is anagrams. But i will not go into that: anything is then possible
and makes the document unreadable and uncipherable to anyone."
They didn't anagram randomly as some decipherers seem to assume. They typically anagramed in consistent ways like this (as just one example):
nøca uøyo adre isth ?
Note the null char for words with an odd number of letters. Another null char (a different one) can be added so that words are of both odd and even length so as not to give the system away:
nøcña uøyoñ adre ñisth ?
This anagram depends on retaining word lengths (a space is a space) but if you assign a character to spaces, even word length can be changed (in fact a char representing the space between words can be assigned instead of the word-length-changing null char and then the actual space char randomly assigned):
nøca ñuø yoñadre ñisth ?
and makes the document unreadable and uncipherable to anyone."
They didn't anagram randomly as some decipherers seem to assume. They typically anagramed in consistent ways like this (as just one example):
nøca uøyo adre isth ?
Note the null char for words with an odd number of letters. Another null char (a different one) can be added so that words are of both odd and even length so as not to give the system away:
nøcña uøyoñ adre ñisth ?
This anagram depends on retaining word lengths (a space is a space) but if you assign a character to spaces, even word length can be changed (in fact a char representing the space between words can be assigned instead of the word-length-changing null char and then the actual space char randomly assigned):
nøca ñuø yoñadre ñisth ?