The Voynich Ninja

(17-11-2022, 05:19 PM)Koen G Wrote: You are not allowed to view links. Register or Login to view.I must say I am very curious about your oncoming thread now, nablator.

I am not going to create a second thread after all, not much to report for the moment. Or maybe in a few years after I check all pages to see how many paragraphs there are where t/k gallows seem to signal a disruption in the alternation (see below) consistently, or if it even happens significantly more often than not.

The problem with large spaces after the removal of gallows is that it is often difficult to decide visually, without doing precise measurements, if their width is closer to the width of the previous character (the unit for normalization) or twice this width, or more, including or not including the gap between characters in the unit... it would be easier to locate all glyphs with a bounding box (as OCRs do) and then do the processing automatically.

There is one black/white dot per EVA character and normalized space, starting with a white dot after each k and a black dot after each t. The idea is that t/k gallows are inserted only when needed to signal an incorrect black/white alternation: two white dots in succession after each k, two black dots after each t:



It is not difficult to find counterexamples, unfortunately. But if gallows are also, sometimes, inserted when there is no absolute need, when the spaces is larger than usual but not clearly as large as two "normal" spaces, or randomly for no reason at all, then there is no way to falsify the hypothesis.


One less ambitious, less problematic project, enough for the week-end , is to check all vords with two gallows whose second gallows is written inside a normal space (the opposite of problematic "floating" gallows). The second gallows should have been inserted there, in an uncomfortably cramped space, only when there is an absolute need for it, i.e. a disruption in the black/white dots alternation, otherwise there is no need for it: a single space would have sufficed.

For example:

f33v.4, f48v.8, f113v.25 tokar(y).

Note: the second gallows could have been missing as in f23r.2 to ar. But this space is quite small, it is actually closer to a half-space, a t or k could not have been inserted there.

A counterexample, with a superfluous second gallows written in a normal space:

f115v.44 ototar

No luck, again.

Are there any significant differences of gallows and their positioning in Currer a and b?

(19-11-2022, 05:00 PM)Scarecrow Wrote: You are not allowed to view links. Register or Login to view.Are there any significant differences of gallows and their positioning in Currer a and b?

Definitely.  Differences between scribes in relation to gallows and their positions is something I'm covering at the conference, although I'm not looking at floating gallows.

I have seen floating (or unattached) gallows only on some Herbal A pages like f27r, f37r, f56v. Other glyphs are frequently unattached too. All B pages and many A pages have a tighter writing.

(19-11-2022, 06:43 PM)tavie Wrote: You are not allowed to view links. Register or Login to view.

(19-11-2022, 05:00 PM)Scarecrow Wrote: You are not allowed to view links. Register or Login to view.Are there any significant differences of gallows and their positioning in Currer a and b?

Definitely.  Differences between scribes in relation to gallows and their positions is something I'm covering at the conference, although I'm not looking at floating gallows.

I can confirm this, even though at the time I only looked at Herbal-A and Herbal-B pages that are interleaved in the MS (25 of each). This result is also statistically meaningful.

(19-11-2022, 05:00 PM)Scarecrow Wrote: You are not allowed to view links. Register or Login to view.Are there any significant differences of gallows and their positioning in Currer a and b?

One difference that may be relevant to this thread is in the ratio of words with gallows to words without gallows.  In Currier A, the probability of a word token in paragraphic text containing a gallows glyph is around 48.5%, while in Currier B it's around 55%.

Those figures are both arguably "close" to 50%, but they're skewed away from it in different directions.  Whatever it is that gallows glyphs "do," it seems they might need to do it just a little less often in Currier A than they do in Currier B.

I'm looking very much forward to tavie's conference paper on gallows positioning and expect we'll have some excellent new fodder for discussion on this topic in a week and a half.

Here is one from Currier B: f31v.8. It looks very much inserted a posteriori.

You are not allowed to view links. Register or Login to view.

(16-11-2022, 03:09 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.Taking the hypothesis that gallows aren't an inherent part of words, but are only added to words, I tried comparing the statistics for all words containing [k] and [t] against the statistics for the same words without them (in paragraphic text only, leaving out labels, etc.).

(16-11-2022, 05:07 PM)nablator Wrote: You are not allowed to view links. Register or Login to view.Yes, gallows can move a lot, with strong preferences for the beginning, next to vowels (counting [c] and [h] as vowels, they function just like [e]). Insertion near or at the end is less frequent.

This way of thinking about gallows placement draws attention to glyph sequences that may be precipitated by the presence of gallows, but are improbable otherwise.  For example:  the rare bigram ye occurs 18 times in the source text, but 321 times "split" by a gallows glyph as y[gallows]e.  The odds are strongly against randomly floating gallows having split almost all of these instances by chance.  If we strip the text of all gallows glyphs, then salt them back in at random positions, only ~23 instances y[gallows]e occur. Random placement is of course not promoted by anyone here as an underlying rule;  as a basis for calculation, however, it can highlight possible laws of gallows-attraction.

If all you have is a hammer... then everything looks like a histogram of character-character correlation frequencies.  The occurrence of trigrams containing a central gallows glyph is summarized below.  Characters preceding a generic gallows glyph are in rows, characters following it are in columns, in descending order of frequency.  Each matrix element represents the count of corresponding trigram pattern matches in IT paragraph text, preceding[gallows]following, divided by the count of trigrams when gallows positions have been globally randomized.  The numerical scale thus gives the factor by which observed frequency deviates from the prediction of a random-placement model:

[attachment=7005]

• The range of values is set by the improbably frequent pattern y[gallows]e.
  
• Another conspicuous pattern with elevated frequency is o[gallows]EVA-vowel.
  
• The frequency of l preceding a gallows glyph is elevated, as is the frequency of S (as Sh) following a gallows glyph.
  
• As noted by @nablator, word-initial gallows are over-represented (entries in the top row -- gallows following a space -- are mostly greater than unity).
  
• As noted by @nablator, word-final gallows are under-represented (entries in the left-most column -- gallows preceding a space -- are mostly less than unity)...
  
• ...but, since the color contrast between 0 and 1 is low, suppressed-frequency deviations are not visible in this graphic.  The matrix element with the lowest value is o[gallows]r, because the plentiful bigram or is split by gallows just once, in okraiin (the only attested example of a gallows glyph directly followed by r).  ~278 instances of o[gallows]r would be expected by chance, giving us a candidate law of gallows-repulsion.
  
• Shown here are fractional deviations from the random model, not absolute counts.  The latter are dominated by o[gallows]e, o[gallows]a, and c[gallows]h as bench.
  

Using Paragraph text from ZL_ivtff_1r.txt, I get higher counts for y[gallows]e: 411 occurrences (probably slightly underestimated).

Occurrences seem to belong to two main families:

• word-initially after a line or image break (e.g. four out of six You are not allowed to view links. Register or Login to view., four out of five You are not allowed to view links. Register or Login to view.). 137 occurrences, 33%;
  
• again word-initially, but after an ordinary word break. 226 occurrences, 55%.
  

Much of the remaining 12% can probably be seen as two words joined together e.g. You are not allowed to view links. Register or Login to view.

Unless I miscounted something, about ~90% of word-initial 'y's are of the form y[gallows] and removing the initial y- typically results in a word with a similar number of tokens. Together with the strong preference for occurring after line or image breaks, I think it's possible that the y-gallows pattern belongs to the transformations discussed by Emma and results from prefixing 'y' to a valid word. So, an explanation for why 'ye' is so rare is that it descends from extreme rarity of word-initial 'e'.

EDIT: though it can be seen in Obelus' graph, it is worth noting that all the occurrences of 'y[gallows]e' involve either 'k' or 't': 'yfe' and 'ype' are absent (with a possible non-paragraph exception in <f67v1.21,@Ro> ykeey.cheo.daiin.ypeeeg.oty, but it could well be 'ypcheg').