13-01-2026, 02:14 PM
(13-01-2026, 01:56 PM)dashstofsk wrote: You are not allowed to view links. Register or Login to view.Positional Preference. Words in the manuscript seems to show a lot of positional preference. For instance if you generate a spline transform of the line positions of words starting with a in quire 20 you will get something like
The frequency of the letter a alone is a bad object of study because it is easily confused with o and y (or even ei), depending on which Scribe wrote it, on how tired/hurried he was, on BEEP!, and on who transcribed that part of the text.
It would be somewhat more meaningful to plot the occurrences of those three letters added together. That should largely reduce that source of noise.
However, again, the frequencies of characters are determined by their occurrence in the most common words; and word frequencies are expected to vary a lot with subject matter, style of text, etc. So, for example, one expects the digraph "rb" to be much more common in a section of an English or Latin book dealing with herbs than in a section about astrology...
All the best, --stolfi