[Talk] Upcoming public lecture on the VM

[davidma]

Ok, people, here's my Univ. of Toronto Medieval Studies lecture! I've got BIG NEWS to report about the potential original structure of the manuscript. Check it out! (new material starts around minute 14) You are not allowed to view links. Register or Login to view.

This was brilliant and very interesting. I still have not understood the maths behind it but can't wait for the paper to be published. It really does feel that between this and the Naibbe cipher that we are finally making some sort of progress. If these were really individual, unbound quadrifolios, I wonder then why the author(s) (which might not be the scribes) felt the need to include a "title" page like f1r. 
I feel that the fact it exists supports the idea that there is some sort of meaning, a quack would have no need for a random page of unreadable text to preface the herbal section, he could just show the individual plants around and it would give him exactly the same result. 
I also wonder how closely related it to You are not allowed to view links. Register or Login to view. and the other 2 of the bifolio. It would be interesting to see if there's some relation or none at all, I would be interested in reading your thoughts on it

Edit: what about quire marks? When do you think were those added?

[LisaFaginDavis]

Thank you so much! The quiremarks would have been added as the manuscript was being organized for the binding, so sometime in the latter part of the fifteenth century.

[Torsten]

Ok, people, here's my Univ. of Toronto Medieval Studies lecture! I've got BIG NEWS to report about the potential original structure of the manuscript. Check it out! (new material starts around minute 14) You are not allowed to view links. Register or Login to view.

Congratulations on your excellent lecture.

The evidence you presented provides compelling support for the conclusion that the Voynich text was executed prior to the binding of the manuscript. In my view, this finding constitutes a significant contribution to our understanding of the manuscript’s production process and adds an important piece to the broader puzzle of the Voynich text.

[LisaFaginDavis]

Thank you so much, Torsten.

[Jorge_Stolfi]

Ok, people, here's my Univ. of Toronto Medieval Studies lecture!

Great lecture, Lisa! (In spite of what you said about Asian languages...   )

I look forward especially to the original order of the Stars section folios.

You mentioned a group who is trying to determine the original order of the bifolios by matching the water stains along the top edge.  If they are using the BL 2014 images, I hope they are aware that they are not at the same resolution.

At least that is the case for the "full size" JPEG images available through the website.  In particular, the images of some fold-outs have lower resolution than the rest.  If not corrected, these differences could ruin the matching of the stains.

I once downloaded a "full size original"  TIFF image but found it to be identical to the "full size" JPEG, with all the JPEG artifacts and all, only much bigger.  But it probably was a single-page image.  Maybe the multi-page TIFFs are truly full size, I don't know.

All the best, --jorge

[Rafal]

As others said - it was a great lecture.

I expected something more "physical" and traditional: a discussion of calfskins, pigments and handwriting styles and instead we got some heavy maths and advanced algorithms. That was surprising but for me it felt okay. It doesn't have to be true but it seems promising and definitely worthy to follow.

Understanding how these algoritms work would probably require a lot of time and effort and I suppose only a few will undertake that effort. But if you show that they work with data that we understand then people will accept it. I can imagine a test in which you:

- take some book, written in a random language
- shuffle its pages
- ask the algorithm to restore the original page order

The important point is that, as you said it, the algoritm is "language agnostic". It doesn't understand words meaning and the bigger sense, it just "sees" the word patterns.

So if I may advise something - do more tests with different texts and see how it works. The problem with Voynich Manuscript is that we don't understandit. If you suggest a different page order based on your calculations then people will say "maybe" and will shrug shoulders   You cannot say now if it's right or wrong. But if you show that your method works with a lot of different works then it will be another story.

[kckluge]

Enjoyed the video of the talk, and very much looking forward to the paper.

* Would it be possible to put a pre-print up on ArXiv?

* You tested the LSA algorithm on the Aberdeen bestiary & Vincent of Beauvais' _Speculum_, but not on any herbal mss. (at least in the context of the work described in the talk) -- wouldn't a mss. of (say) Dioscorides be a better baseline for comparison with the herbal quires?

*  With regard to the XRF work -- is there any possibility or work-in-progress of narrowing down the source of the pigments (in particular, the azurite) by comparing the distribution of the trace elements with XRF results from other mss. or samples from mines known to have served as historical sources? There seem to be other groups looking at mss. with trace amounts of barium in their azurite (You are not allowed to view links. Register or Login to view., for example).

* Anikó Bezur suggested back in 2016 that "...Raman spectroscopy revealed the presence of quartz crystals in the manuscript’s red pigment...further analysis of the quartz crystals might indicate where the sand came from, which could provide evidence of the manuscript’s geographic origins." (You are not allowed to view links. Register or Login to view.) -- is it viable to/has there been any work/progress on following up on this thread?

* I was very, very surprised by the statement @ ~25m17, "Although linguists will tell you that they're not really different enough to be called different languages, hence I'm putting them in in quotes." 

While Bowern and Lindemann discuss (in Section 3.3 of "The Linguistics of the Voynich Manuscript") the poor overlap between the 10 most common words in the A and B languages (apparently grouping Herbal B with Bio B into a single "B language"), they waffle by saying, "While there is some overlap, the most common vocabulary items of Voynich A and Voynich B are substantially different....They might be the result of different encoding processes, or they might [emphasis added] represent different underlying natural languages."

The problem isn't simply that they don't provide any example of a natural language case where difference in topic produces this level of lack of overlap in most-common words -- the problem is that they don't note the poor overlap between the most-common words in the Herbal A and Herbal B folios in the first herbal quires: (Sorry for the Currier.)

Rank:      1       2        3        4      5       6      7       8       9           10
HerbB:    8AM   SC89    OR    AR    AM    8AR   89     S89   4OFC89  ZC89
HerbA:    8AM   SOE    SOR    89    S9     2      ZOE   Q9     8AN       ZO


Hard to see how that could be attributed to differences in topic or author word choices. I'm willing to be convinced by evidence from some corpus of natural language texts that the lack of overlap there doesn't very strongly imply that Herbal-A and Herbal-B are different languages/cipher keys/whatever, but until I see it I'm definitely mentally living in Missouri (the "Show Me" state for non-US readers)...

[LisaFaginDavis]

As others said - it was a great lecture.

I expected something more "physical" and traditional: a discussion of calfskins, pigments and handwriting styles and instead we got some heavy maths and advanced algorithms. That was surprising but for me it felt okay. It doesn't have to be true but it seems promising and definitely worthy to follow.

Understanding how these algoritms work would probably require a lot of time and effort and I suppose only a few will undertake that effort. But if you show that they work with data that we understand then people will accept it. I can imagine a test in which you:

- take some book, written in a random language
- shuffle its pages
- ask the algorithm to restore the original page order

The important point is that, as you said it, the algoritm is "language agnostic". It doesn't understand words meaning and the bigger sense, it just "sees" the word patterns.

So if I may advise something - do more tests with different texts and see how it works. The problem with Voynich Manuscript is that we don't understandit. If you suggest a different page order based on your calculations then people will say "maybe" and will shrug shoulders   You cannot say now if it's right or wrong. But if you show that your method works with a lot of different works then it will be another story.

Thank you so much! I didn't have time in my lecture to go into details about the other texts we tested - but we did indeed use several different manuscripts to test the method on other texts. This will be laid out in our article.

[Jorge_Stolfi]

With regard to the XRF work -- is there any possibility or work-in-progress of narrowing down the source of the pigments (in particular, the azurite) by comparing the distribution of the trace elements with XRF results from other mss. or samples from mines known to have served as historical sources?

XRF only determines the elements present in the sample.  Thus for azurite and malachite it basically says "copper".  It can detect minor elements, but I doubt that it can use them to identify the source of the pigment. It is not that precise, and the amounts will have been distorted by the preparation of the pigment and contamination over all those years.

I think one could determine the source of the pigment by isolating some crystals of azurite, clean from other contaminants, and determining the elements and isotopes in them.  But I don't think that labs like McCrone have equipment that can do that.

Such investigation could perhaps give useful information, but pigments like azurite were traded all over the world. Thus the source of the pigment would not tell for sure where the paint was applied.

Much more informative, in my opinion, would be to locate the pages that Baresch apparently sent to Kircher.  I bet that they were not yet painted... 

All the best, --jorge

[ReneZ]

Much more informative, in my opinion, would be to locate the pages that Baresch apparently sent to Kircher.  I bet that they were not yet painted...

Oddly enough, the single page found by Fabrizio Salani, which he claims may be one of these pages, is painted, but using modern pigments.