Why and how the text could be Bavarian

[Jorge_Stolfi]

I think that the first word of a page can’t really be a continuation of a word from a previous page. If anything that initial glyph has to set the page start somehow. There are other quirks like split gallows that do something specific to the beginning of a page along with the weird paragraph first line use of f and p.

While some page-initial glyphs are extra fancy (like on You are not allowed to view links. Register or Login to view. and f42v), most of them seem to be no more fancier than other paragraph-initial glyphs. 

So I don't think that the start of a page is a semantically significant place.  I think that the extra-fancy gallows are just decoration provided by the Scribe. 

On the other hand, I expect that the start of a paragraph will be semantically different from other places in the text.  And I don't include here the use of puffs (p and f gallows) in place of other glyphs or glyph combinations, which I take to be a purely decorative decision by the Scribe.

The most likely contents of the Herbal section is indeed a herbal, even if "fake" like the You are not allowed to view links. Register or Login to view.   Then each paragraph in a page is likely about one useful part of the respective plant: fruit, bark, leaves, root, etc.  Then we can expect each paragraph to have the same loose structure. Like, first the name of the plant or part, then a list of uses, each possibly with dosage, method of preparation, etc.  Then maybe end with the type of location where the plant grows, and perhaps method of harvesting (see the entry for "mandrake" in any good medieval herbal).

One consequence of this hypothesis is that the contents of each page should be independent of the other pages, apart form the words and phrases that belong to the "herbal" topic in general (like "herb", "tea", "cures", etc.).

Under this hypothesis, the pages should have no "natural" order.  Needless to say, if one orders the pages by some statistical similarity criterion, then their statstics will seem to vary gradually along the sequence...

All the best,--stolfi