I had an interesting note from long-time Voynich Manuscript researcher Dana Scott about the Weldon Ciphers (thanks!). He noticed that (what Chris Demwell idiosyncratically called) the “fish” and “dog” glyphs in fact appear to be a linked pair:-
So “fish” actually = “fish head”, while “dog” actually = “fish tail”. But what that means is still anybody’s guess.
Another interesting angle is, as the Pressing Refresh blogger points out, the moderate similarity with a game played in the same D. B. Weldon library in 2011. This was based around a fictional “Captain Smith” and included all manner of harmonic-themed cookie crumbs for the players to find. There’s much more on the Intangible Harmonics website, according to which it all ended up like this:
“Our players determined that Smith’s words were from an obscure poem called Tecumseh, Or The Warrior of the West by John Richardson. The players then somehow figured out that the words on the scrap of paper were in the Shawnee language, the same language spoken by the great Tecumseh and his brother Tenskwatawa. The words were actually a library location code.“
So, that’s one set of people who might possibly be involved, perhaps as a kind of karmic payback for (or follow-on to) Captain Smith’s historic adventures. Who can tell?
The Pressing Refresh blogger also suspects it might be a certain cryptographically-focussed assistant professor at Western University, which would also make a lot of sense.
My own best guess is that it is a cryptic literary cipher (for I’d be a bit surprised if a crypto professor would allow letters to scroll off the bottom of the page when inserting different-sized pictures: and pink feathers too?) by a writer who previously had his/her own blog hosted on the blog.ca domain. I managed to find a “Pale Writer” blog active there from 2006-ish, so – in the absence of anything better – that’s today’s best guess. Whoever that is (and perhaps you know who that is, because I certainly don’t).
Just so you know!
I noticed the fish thing a week back and it looks like the maker meant something by splitting the icon but I can’t figure out what. There is really no other reason why he would not have just used the whole icon unless it is a clue.
By making the font for note one (the leaf) I can switch back and forth between symbols and the letters I used to represent them, (A – B and a – u (47 symbols)). This has allowed me to run I.C., Kasiski, frequency, etc. but nothing is giving any clues although the frequency dist. is not flat and suggests some non-random pattern.
I am going to add the other symbols in the additional notes and re-run some stats but again, the fact that all have non-repeating symbols in the first 17 is peculiar and suggests that either the notes are very similar or a hoax.
Shark? Trolling? Sharks have to keep moving instead of ‘finning’ in order to breathe. So ‘fin – ish’ instead of ‘fish’?
If the two split icons are on either side of a section of text they may function like quotation marks or the ancient egyptian cartouche by telling information about the enclosed text.
I have just looked at several pages of the Weldon cipher and I have come to the conclusion that there is only one “fish tail / fish head” pair on each page and that they always appear in the order “fish tail” “fish head”. Could they be some kind of function? (Like the “on/off” switch in Sandra and Woo). Or do the symbols enclosed by them mean something special?
Russell: certainly could be. I’ve only done fairly lightweight cryptanalysis on the Weldon Ciphers so far (because it seemed that getting the transcriptions properly locked down was the right first step), so I’ll be looking at this kind of thing in more detail before very long.
PS: don’t forget that we also have the bird/ra (hieroglyph) pair, I haven’t yet looked for patterns with them! 🙂
There are more pairs like dotsquare/dotcircle. Maybe paired signs belong to each other.
Nick, I’ll leave a comment which seems trivial and obvious — and hope I get an education.
Most of the techniques I’ve learned are “classical” and statistical. They are designed and work well when the *sequence* of an alphabet are known. What’s getting me puzzled again and again is the lack of a known sequence to effectively work with these techniques in this study.
I’ve tried approaches that had some success with Beale#2 (treating the word number as a virtual symbol, and knowing English language sequence of letters)… not as much success this time (nor final success on the other Beale’s either). Here, Vowel-Following-Consonant matrix decomposition is impacted by so many symbols being used infrequently — I am not trusting that all three version use the same ciphertext mapping.
Also, I’m lacking a working crib (shades of #408) an alternative toehold into the Homophone barricade (more than a few failures there, as the number of symbols within each note is just so many to almost be a One-Time-Pad basis).
Right now, I’m not seeing how intra-symbol-count distances between all possible pairs of bounding symbols will give me a clue what next to do (easy to code, difficult to figure what to do with that gathered data yet).
Other than hoping the composer was utterly lazy at mapping font character positions to actual letters (perhaps Escher’s hope as well?), I will drift back into lurk mode and hope the Encyclopedia Girl provides a Rosetta Stone hint into the Weldon hieroglyphics, I might as well just stare at Linear A & B again for inspiration.
Please don’t misunderstand. I love some struggles!! And I admit I need some help now. Learning from the masters and more experienced is better than throwing genetic style software randomizing symbol to letter mappings in dart board fashion in utter hope/frustration (which will almost be worthless if any form of Translational remapping was used). I don’t want stupid luck; I’d like to understand technique.
Many thanks!
Dennis: in my experience, the thing about historical ciphers is that you have to go through a process by which you become sure that you’re reading the shapes on the page correctly before you launch into a lot of technical cryptanalysis. For instance, the fact that there appear to be different length variants of the same ciphers on different notes seems to point away from transposition ciphers: and a new note has been found in the last few days with what seems to be an insertion near the end, which would also point strongly away from transposition ciphers. Lots to get right…