Analysis of patterns at the very start and end of lines, per line type

[oaken]

I don’t like throwing out baseless theories or overcomplicating things, but I can’t help noticing that the characters most common at the beginnings of lines (except the first lines of paragraphs) are also the ones that most resemble Arabic numerals in the Voynich.

So the first letter could be a key to use in encryption/decryption algorithm for the rest of the line?  Like the shift of a Caesar cipher, or the column of a table of more arbitrary encryption substitution ciphers?
[/quote]

I can't remember who posted it but in another recent thread, a user posted a page where it appeared some of the line-start letters of some lines seemed to have been written around existing letters (something I understand earlier researchers have rejected on the whole for the vms), but I did notice if I covered the rest of the lines it looked like a column of Arabic numerals to my untrained modern eyes.

However, there is the problem that any position-dependent cipher would almost surely encode each occurrence of the same word type as two or more distinct words.  That would flatten out the frequency x rank plot ("Zipf plot") at the high frequency end -- which does not seem to be the case.

Do you think it would make this stat more like vms if these line start numbers somehow functioned with the also anomalous paragraph start lines to form a key for the paragraph?
(I understand such a solution has implications for the reading of labelese, and I tend to agree that the vms seems an unlikely cypher in the traditional sense, although cyphering know-how could have been part of the generation process).