Relation between word types and vocabulary density.

[Jorge_Stolfi]

I will try to explain in my own words what I think is happening here. If we assume bifolios are sequential and not supposed to be nested (singulions vs quire structure), then three seemingly different approaches point to the same order.

I don't have clear feelings about the order of the bifolios of quire 20, except that the bottom half of You are not allowed to view links. Register or Login to view. must be some "postface" for the Stars section (if not for the whole book), and f116v, which originally had only one parag, must be the last page.   Thus

• If the bifolios were supposed to be nested, f103+f116 should be the outermost one;
  
• If they were supposed to be "singulions", then f103_f116 should be the last one.
  

Anyway, note that the page similarity criterion may recover the original reading sequence, but not the direction.  So, with no other consideration, all four proposed orders would be just as likely as their reversals.

Was it considered that some bifolios could have been folded the other way?  That is not an option for f103+f116, but maybe for the other ones?

There is another detail that may have to be accounted for.  On pages f108v, f111r, and You are not allowed to view links. Register or Login to view. (currently the central bifolio, after the loss of f109+f110), there are three large blocks of text where the Scribe merged 10-12 parags without any clear parag breaks (but still drew the respective stars in the margin).  

One explanation I can think of is that the Scribe started doing that -- maybe for fear of running out of vellum, maybe because he did not think that parags were important; but then was told by the Author to properly break the parags.  How/when would that have happened, exactly, in the new arrangement? 

And the bifolio f105+f114 is special too.  Besides being cut in such a way that the bottom of f105 seems to be missing up to 2cm of vellum, the initial fancy gallows of You are not allowed to view links. Register or Login to view. is fancier than those of the other pages.  Maybe f105+f114 was the first singulion?

(That bifolio is very peculiar also within the Global Restoration theory.  But I will leave it at that.)

On the other hand, one must be wary of interpreting those "statistical" orderings as the scribing or writing order, or the intended reading order.  One old book that shall remain nameless but seems to have a structure very similar to Quire 20 is a materia medica that classified remedies into three categories: tonics that can be taken at will, remedies that should be taken only to cure specific conditions, and toxic remedies that should be used only as a last resort, in the smallest possible amounts, with close monitoring by a doctor.  Each category has a specific vocabulary.  Most entries of the first category have the words "extended use", and the benefits are often repeated, like "extend life'.  Entries of the second category do not mention "extended use", and list diseases instead of long-term benefits.  Etc.  

Now, in some editions of that old book, the recipes are sorted by category.  But in other editions they are sorted by other criteria, like minerals/herb/animal/insects/trees/grains.  In the first case, comparison of vocabularies between pages might recover the intended reading order.  But in the second case that same criterion may scramble it...

If the entries of Quire 20 also had similar categories with specific vocabularies, they may or may not have been sorted by category.   Moreover, it is possible that the Author copied some ur-book where the recipes were sorted by topic, but then the order got scrambled when the Author copied the original notes to a final draft, and/or also when the Scribe copied the draft to vellum; in both cases scrambling the pages of the source, ignoring its page breaks, and freely re-paginating the copy.

Finally, I would rely on "singulion" hypothesis to propose that neither the Scribe nor (probably) the Author were aware that books are bound by quires.  They may have thought that books were stacks of singulions sewn together. (Weren't the first codices, say ca 100 AD, bound like that?) That would explain why the Scribe wrote the entire book as a series of singulions (and why the Author did not stop him).

All the best, --stolfi