Re: VMs: repetitions repetitions

[Koontz John E <John.Koontz@xxxxxxxxxxxx>]

On Fri, 13 Aug 2004 elvogt@xxxxxxxxxxx wrote:
> Word length distributions could be explained if the VM script is actually a
> Roman numbering system in disguise. (I think somebody already pointed out that
> VM word length matches the length of evenly distributed Roman numbers quite
> well -- is that correct?)
>
> So you could have something like
>
> IX.VII.IV.III
>
> to indicate "book 9, chapter 7, verse 4, word 3".

A difficulty with this approach in highly inflected languages is that you
either have to ignore inflection, not always feasible, or you have to find
properly inflected words in the source text, which can be difficult.  I
think this explains why the Italian coding systems that D'Imperio
exemplifies pair code-book-based schemes for roots with auxiliary material
to indicate the inflection of the form.  In these cases the source book
was apparently a code-specific one, but that wouldn't be strictly
necessary.

> Indeed, if this is the method used, it's going to be next to impossible to
> decipher.

Unless the book was a common edition of a common work, post printing
(i.e., with standardized page contents).  Of course, the earlier the
preparation is, the less likely this is!

> One method to test it, though, might be paragraph initial and terminal
> words.  Perhaps the frequencies of these words in the VM can be shown to
> coincide with similar words in real life plaintext sentences.

With this scheme a given word might be encoded at different times with
different indices for different occurrences in the key text, though ease
of preparation would suggest keeping this to a minimum!

______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list