Re: Letters coded by groups of symbols...?

[Nick Pelling <incoming@xxxxxxxxxxxxxxxxx>]

Hi everyone,

> -       The first 8-10 characters of each page often don't fit

Given that my hypothesis is that gallows get converted to simpler 
characters (like "o" or "c"), I decided to evaluate the possibility that 
one of the frequent gallows combinations - "o + gallows" - is actually some 
kind of reference into characters 2..9 at the top of the page (which often 
appear not to match the pairing structure of the rest of the text).

ie, on the herbal page f21r (which Dana identifies as Rupturewort, and 
which is in Dioscorides as Empetron, I think), the first line is...

   pchor.oeeockhy.ofychey.ypchey.qopcheody.otaiin.chan-

.....whose first nine characters are...

        1       p [embellished]
        2       ch
        3       o
        4       r
        5       o
        6       ee
        7       o
        8       ckh
        9       y

If this is the case, then all we'd need to decode "o + gallows" pairs for 
this page is a table of correspondences between the order of characters 
2..9 and the 8 gallows characters. Putting in Steve Ekwall's folding key 
gallows order (for want of anything more definitive) produces a table like 
this:-

        #       f21r    KEY ORDER
        1       p       -
        2       ch      t
        3       o       cth
        4       r       k
        5       o       f
        6       ee      cph
        7       o       ckh
        8       ckh     cfh
        9       y       p

So: by this table, the EVA "o + f" on the first line above would code for "o".

However, looking at the surrounding text, this would be unlikely as "o" 
would not convert into a pair with either the previous or the following 
character (both of which are "y") - and "oy" isn't a combination that looks 
likely to be a pair. So: this probably isn't the correct order.

Let's try a more rigorous approach, by looking at all the "o + gallows" 
pairs on this single page and their immediate (EVA) contexts:-

"o + f"
            y.ofy
"o + k"
              okaiin
           ol.okeoaiin
            y.oky
"o + p"
           al.opsh
"o + t"
           ol.oteos
          ain.otol
          iin.otol.ch
          shy.otor
         aiin.oty
            y.otaiin
          eey.otaiin-
         cthy.otyky.
"o + cph"
        sheol.ocphal

What does all this tell us about this page?
(1) "o + f" codes for a character that precedes "-y"
(2) "o + k" codes for a singleton character that sometimes precedes "-y"
(3) "o + p" codes for a singleton character
(4) "o + t" codes for a singleton character that sometimes precedes "-y"
(5) "o + cph"  codes for a singleton character

There are 5 "o + gallows" combinations used in the page and 5 non-"o" 
characters listed in the header: "ch", "r", "ee", "ckh", and "y".  [ Quite 
what it means for a gallows character to be replaced by another gallows 
character we'll leave as an open question for now. :-/ ]

However, "r" doesn't appear to pair up satisfactorily in any of the cases 
listed above: so we're probably back to the drawing board. :-(

So: this doesn't look like it's the code we're looking for - but it was 
still a pretty good try. :-)

All the same, "o + gallows" here are all likely to code for free-standing 
singletons: it will be interesting to see if this holds as true for other 
pages in the VMS...

Cheers, .....Nick Pelling.....