(15-12-2019, 11:23 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.The strange bit is that this behaviour is also observed if 'chd' is somewhere in the middle of the word and not at the end. That is, if I saw that correctly.
Hi Rene,
the numbers I computed are for all occurrences of chd (anywhere in a word). Also in A, it typically occurs as -chdy: this is the case for 7 of the 14 end-of-line occurrences. The other 7 cases are:
chd
kchdal
ytchdam
qotchd
ykchdg
tolchd
opchdard
Since the numbers are so small, I find it difficult to evaluate more specific patters. In my opinion, even if the position of chd is not strictly word-final, it could still be some kind of 'contraction' as suggested by Nick.
The [chd]-phenomenon appears especially prevalent in the large-plant pages, where for the A-pages I counted:
8 line-end
10 before image
10 mid-line
Which means an 18 to 10 ratio of blocked vs free. But this might simply be the case because on these pages there is much more interference from the images in general.
Also, in the small-plant pages, most occurrences are labels, but not all of those are constrained by lack of space...
I deliberately avoided the Pharma pages. Basically, because I didn't want the stats to get tangled with label behaviour, I specifically focused on Herbal-A pages.
The next thing to search for would be other bigraphs / trigraphs that occur noticeably more often in span-terminal (i.e. "immediately before an interruption or at the end of a line") positions than in non-span-terminal positions in Herbal-A pages, in the same way that chd seems to do. I've tried a fair few manually without luck, but perhaps cunning scripts will succeed where the eye fails. :-)
Alternatively, it may be revealing to use the same cunning script to try to find bigraphs / trigraphs that occur noticeably less often in span-terminal posiitions. This may reveal parts of Voynichese's writing system that get abbreviated in ways that are perhaps not as directly visible as the (putative) chod--chd contraction.
One note: "line-final" does not necessarily mean "no space available for writing". In fact, in most cases in herbal folios it would not mean that.
Anton: there is also the complicating factor that the right-hand edge is actually not very ragged, which implies that all line-terminal words apart from paragraph ends may often have had to be compressed to make them fit nicely.
So perhaps "span-terminal" should be restricted to a definition more like "immediately before interrupting drawings OR (line-terminal AND NOT paragraph-terminal)".
That's my impression as well, they seem to somewhat respect an invisible margin. It's not perfectly justified, but not too irregular either. So the possible contraction at line end would rather be applied for visual effect.
Space is not an issue in the first place on any Herbal folio.
Rather than considering a virtual right margin in the final manuscript, I would prefer to consider a true margin in a source document, for example a draft.
After looking at some books and reading some texts, I can't find any rule to separate. Since even names like Maria are separated somewhere, I should judge not only the last word of a line, but also the first in the following line.
What if you simply list all line end + blocked by image tokens in Herbal A and compare glyph / bigram frequency to that of the whole section? Or would this be too rough?
Koen: that's exactly the right starting point, though you'd probably need trigraphs etcto avoid going too fine with EVA strokes, e.g. octh.