Here are some piquant canapes to twingle your Voynich tastebuds, a bit like “Space Dust for researchers”.
(1) Gerry Kennedy has discovered that a deathmask of Wilfrid Voynich was taken, and that it still exists.
(2) Jackie Speel tells me that “in 1916 Wilfrid Voynich was involved in a friendly law case with the Lincoln Cathedral authorities over the ownership of one of their books he had acquired in good faith from another American dealer (The Times May 11, 1916 pg 4 refers). In the event he donated the book back to the cathedral.”
(3) Jackie also notes that Wilfrid Voynich’s British Museum (i.e. what is now the British Library) “ticket/renewal details still survive – he first joined on 19 October 1895 and the ticket was last renewed 29 November 1907.”
(4) Diane O’Donovan points out that “Chinese inks are high carbon”.
(5) While Ludi Price concurs that f1r’s “glyph 3 does look uncannily like yuan, the [Chinese] character for ‘first'”, the key problem with VMs Chinese theories is that “(spoken) Chinese in the 14th/15th century was completely different to modern Mandarin Chinese. It had more tones, more glutteral stops, and was more akin to modern day Cantonese than it is today. People who want to tackle a VMS as Chinese theory would need to approach it from a Classical Chinese standpoint, not a modern one.”
(6) Henry Berg has just published his own breathtakingly syncretic Voynich theory, and hacked it on as a link to the bottom of the Wikipedia page (shame on him). It’s a heady mix of Francis Bacon, William Shakespeare (Hamlet in particular), Athanasius Kircher, Isaac Voss and even Shugborough Hall (no, I kid you not), with 17th century conspiracies and disinformation aplenty. Great fun for the the next Voynich pub meet: but little genuine chance of being a workable hypothesis, alas.
(7) On the diametrically opposite side of the color wheel, here’s a reasonably balanced (but, even so, frequently wrong) view of the VMs’ ciphertext, courtesy of Sravana Reddy and Kevin Knight, hot off the presses. Enjoy!
http://imageshack.us/photo/my-images/860/asdfoj.jpg
Something unseen about VMS ?
Rather taken by “glutteral”…
The Reddy and Knight paper is plainly the most significant item here. My first reaction is that this is good work, but mainly serves to confirm negative results and facts that we already knew. I would summarise it as follows.
1, 2 Introductory
3.1 Not original work.
3.2 If you replace a gallows character with a non-gallows character you get more entropy (less structure). True but not new (the gallows are a substitution class in all proposed word grammars).
3.3 Not original work.
4.1 Not original work.
4.2 Not original work.
4.3 The MDL algorithm confirms the prefix-midfix-suffix idea. Interesting but not new information.
5.1 The occurrence of a word is only very weakly predicted by the previous word or pair of words. True but not new.
5.2 The final and penultimate characters of previous words are more predictive than entire words as in 5.1. I believe this to be a new result.
5.3 Scrambling sequences of words does not decrease bigram frequency. Therefore no long distance correlations between pairs of words existed in the first place. I believe this to be a new result and an interesting technique.
6.1 There is a weak tendency for word frequency within a page to differ from word frequency in the entire corpus. True but not new.
6.2 Character frequencies are different in initial, medial and final position. True but far from new.
7.1 A specified measure of similarity between pages is strongest between adjacent pages. I believe this may be new.
7.2 Not original work.
8 Not original work.
Philip: from where I’m sitting, I suspect the authors are too subtly wedded to a whole set of just-plain-wrong presumptions (that there is a one-to-one letter mapping between ciphertext and plaintext, that the same word repeated in the plaintext would necessarily repeat in the ciphertext, that it is either normal or an abjad but not inbetween, that any punctuation would be essentially modern, that the pages are probably mostly in their original order, inadequate oversight of the codicological and palaeographical data, etc) that skew their range of questions to test so significantly as to make them practically useless. For example, the tests in 7.1 between pages would yield completely different results across recto-verso boundaries as compared with verso-recto boundaries (except in those cases where a verso-recto page was at the centre of an original quire). Which is not to say that just about everybody else has made the same set of presumptions (because they have), but rather that it’s somewhat dismal that people are still falling into the same basic traps after all these years. Oh well! 🙁
What intrigued me was that the work was partially funded by the NSF! I spend a lot of time writing proposals to NSF, and wonder whether I’ve picked the wrong topics (e.g. cyber physical systems to help with disaster relief), since it’s so bloody hard to get funded.
This will also be breathlessly mundane – but Arabic inks and inks used till roughly the 12th-13thC in the west are also high carbon (as if we didn’t know).
In mentioning chinese inks, for a work only known in the west and dated to the 15thC, I was thinking of how much territory happened to be occupied by the Mongols in the early 15thC.
I could imagine someone in touch with the Greek or Latin world – like Montecorvino or one of the Alans – might have tried to have his own alphabet the one finally adopted for recording Mongolian. No evidence though.
Rather taken by “glutteral”…
As a linguistics student, so was I. I’m guessing it’s a slip for “glottal” (the glottal stop is how Cockneys pronounce the “t” in “bottle” and a common consonant in Hawaiian, but I don’t know its status in Chinese).
Hi,
how can I get in contact with an expert in the Z340?
I’ve got some ideas.
A current (March 2012) link to the Reddy-Knight paper, now available as a pdf.
http://www.aclweb.org/anthology/W/W11/W11-1511.pdf
One look at the bibliography was enough to deter. But then I looked at entry for codicology..
nuff said.
I’ve been out of the Voynich loop for the past few months, and was (pleasantly) surprised to see I had been quoted here.
And yes, I meant glottal stop. Sorry. How embarrassing.
Glottal stops are used in modern Cantonese and some other Chinese dialects. It not used in modern standard Chinese (i.e. Putonghua). So if you were a supporter of the ‘Voynich-as-Chinese’ idea (which I’m not), and that the VMS writing system was a form of Chinese transliteration (like Pinyin), you’d have to keep in mind that the underlying language of the text would have been expressed differently to the way it would be today.
Keep up the good work on the blog, Nick! I do so enjoy reading your posts, fellow Surrey-ite 😉