[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Re: VMs: A very important discovery!!

On Tue, 28 Sep 2004, jan wrote:
> There are the sections in the VM, where the SAME gallows follow each
> other - in the same paragraph and with the other characters in between
> but without different gallows in between. What would be the need for
> that (one gallows would normally do)?

That's another argument for not seeing the gallows as keys, but somone
could always argue that in the absence of a leading gallows a standard
value is to be assumed, while in the presence of a superfluous one one is
facing misdirection on the part of the encrypter.  This last is
Ridderstad's argument, as I understand it.  One could also argue a policy
of entering a "selector" whenever resuming encypherment after a pause.

> Or maybe the particular gallows modify say only the following character?

Another possibility, certainly, though in that case we have a rather
different scheme of encoding.  You could think of it as a five table
Ridderstad scheme, with the absence of a gallows indicating the fifth
table.  Actually, in hypothesizing (not very successfully) that the
gallows might be part of a scheme of encoding secondary consonants in a
cluster, e.g., r and l, I was suggesting something like this - a gallows
character with local effect (+r, +l, etc.).

> Second point: what about split gallows?  Here maybe perfectly reasonable
> function:  they may "modify" only the text in between the vertical
> "beams".

Yes, though this approach only works with "local-effect" gallows.  If
gallows only occurred at the end of a word (and there were only one), they
might indicate, say, superscriptingm the convention modern indicator of
"reduced letter inclusion" or abbreviation, as in y(raised e, t. etc)
for the, that, etc., or 1er, 2me, ... or 1st, 2nd, ...

Unfortunately for that notion there are four gallows characters.  This
also plays hob with the idea that they might represent r and l, actually,
unless one factors off the second or flourish part.  But that produces the
oddity that r and l are only followed by two of the flourishes, unlike e
and i (sequences), which take a full set of five or six.

Thinking in terms of a letter scheme, however, whether for clear text or
encryption, there's no reason why some letters might not have one
graphical logic and others another, with no particular phonetic logic.
This appears in the case of something like Arabic, where there is a good
deal of similarity in "orthosyntactic behavior" of letters within the
several classes of character types, but there *are* several classes of
character types, and membership in one or the other doesn't imply
phonological similarity to other letters in the same class.

This is of course quite different from the approach in Tolkien's Tengwar,
where similar graphic behavior is strongly correlated with phonological
behavior.  Of course, Tolkien was a product of (early) modern linguistic

> The Ridderstad's approach is truly the novel idea. We already noticed
> the gallows might have different function than to represent just the
> characters - so here is the one option.

She's quite specific as to the nature of the scheme, but isn't it
essentially the idea of "encryption control character" that's been
floating around?  I suppose one difference is that the scheme of
encryption is implicit, rather than depending on, say, the next ten
characters following the gallows, or everything between one and another,
or whatever people were hypothesizing.

A third point you might raise is the function of embedding gallows within
benches.  Does that yield an extended set of control characters, or do we
follow Ridderstad in supposing, essentially, an arbitrary ligature of the
gallows with something either following, or, alternatively, preceding.

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