How perfect is Voynichese?

[ololololo]

Can you read the following? How quickly? Genuine question, I don't really understand whether reading these anagrams is a skill that can be learned. Please don't spoil it for others, would be nice to collect a few opinions.

CENO OUY AEKM A AEIKMST EENV A EILMPS AEHPRS ACN EGT ADEHRR ADN ADEHRR OT ADER, EEHORVW IHTW EMOS ACEIPRT ABEMY HIST ILLW KORW

I can read it.  Pretty easily, in fact.  I didn't time myself, but my impression is that it took me about three times as long as it would have taken me to read non-anagrammed text.

Can I tell the answer?

[Grove]

5 letter words and less are extremely easy. Once you have the context, even errors are easily worked out. Not quite as easy as words written backwards, but close.

[ololololo]

5 letter words and less are extremely easy. Once you have the context, even errors are easily worked out. Not quite as easy as words written backwards, but close.

In the case of longer words, you can find a solution by moving certain letters or parts of the word to the desired location. Example: CONCAEDES - SNEEDCCA. I can just leave the prefix and ending the same, and it will be easier for you to guess the word: CON ACDE ES or CONACDEES. In the case of a cipher, these parts of the word can be encrypted, for example, as individual characters (a person familiar with medieval literature will have no difficulty encrypting a couple of prefixes that were often shortened).

[Grove]

The longer words only take a fraction longer than those less than 6 letters, but those less than six are pretty much immediately understood.

[ololololo]

The longer words only take a fraction longer than those less than 6 letters, but those less than six are pretty much immediately understood.

All right

[oshfdk]

I can read it.  Pretty easily, in fact.  I didn't time myself, but my impression is that it took me about three times as long as it would have taken me to read non-anagrammed text.

This means that sorting letters in words can be a viable method of text obfuscation, which still leaves the message readable. Thanks!

[Jorge_Stolfi]

CENO OUY AEKM A AEIKMST EENV A EILMPS AEHPRS ACN EGT ADEHRR ADN ADEHRR OT ADER, EEHORVW IHTW EMOS ACEIPRT ABEMY HIST ILLW KORW

I recall a claim that, if one scrambles the letters of every word in an English text, keeping only the first and last one fixed, people will still read most of the words at almost normal speed -- and often not even notice that a word is spelled worng.

Your text only takes the scrambling a little bit further, so it is a little bit harder to read...

All the best, --stolfi

[ololololo]

CENO OUY AEKM A AEIKMST EENV A EILMPS AEHPRS ACN EGT ADEHRR ADN ADEHRR OT ADER, EEHORVW IHTW EMOS ACEIPRT ABEMY HIST ILLW KORW

I recall a claim that, if one scrambles the letters of every word in an English text, keeping only the first and last one fixed, people will still read most of the words at almost normal speed -- and often not even notice that a word is spelled worng.

Your text only takes the scrambling a little bit further, so it is a little bit harder to read...

All the best, --stolfi

Exactly
It is worth noting that if you combine anagrams with substitution, you will get a fairly secure cipher, no worse than a one-time pad... 
Anagrams blur the statistics, making it impossible to detect substitution. If letters are encrypted with numbers, and numbers are encrypted with other symbols, the resulting anagrams will be completely unanalyzable.
There are other tricks to make the simplest anagrams more difficult. One of them is to blur the boundaries between letters in the word itself (daiin - d-a-i-i-n, d-a-iin, da-iin, d-ai-in, etc.). 
In any case, I believe that the text would have been transcribed manually, and the computer would not have been able to help us.

[oshfdk]

Anagrams blur the statistics, making it impossible to detect substitution. If letters are encrypted with numbers, and numbers are encrypted with other symbols, the resulting anagrams will be completely unanalyzable.

Anagrams over simple substitution won't blur character statistics, so it's still relatively easy to break, if the language is known.

[ololololo]

Anagrams blur the statistics, making it impossible to detect substitution. If letters are encrypted with numbers, and numbers are encrypted with other symbols, the resulting anagrams will be completely unanalyzable.

Anagrams over simple substitution won't blur character statistics, so it's still relatively easy to break, if the language is known.

Yes, but not always. It depends on what substitution we apply. In the case of VMS, for example, knowing the language won't help us much.