New clue to Voynich manuscript mystery(theguardian.com)
theguardian.com
New clue to Voynich manuscript mystery
http://www.theguardian.com/books/2014/feb/07/new-clue-voynich-manuscript-mystery
23 comments
If it is Nahuatl, it should be possible to decipher a lot of it pretty quickly. Nahuatl has very regular morphology which will give repeating patterns that can be used to decipher the letter, assuming they're phonetic. For instance most nouns end in -tl so we should see many of the words in the manuscript ending in a symbol or two symbols which we can assume is tl.
Maybe someone can apply this model: http://pages.cs.wisc.edu/~bsnyder/papers/bsnyder_acl2010.pdf
Maybe someone can apply this model: http://pages.cs.wisc.edu/~bsnyder/papers/bsnyder_acl2010.pdf
I am also interested to see what comes of this, but it's been pointed out that is paper is somewhat suspect with seemingly little input from Nahuatl language experts or prior literature on the Voynich manuscript.
Here is a refutation by one of the leading "hoax" proponents: http://hydeandrugg.wordpress.com/2014/02/06/hoaxing-the-voyn...
In it, he explains the prevalence of half-baked origin theories as a function of several psychological and probabilistic phenomena:
* Pareidolia * Confirmation bias * The birthday problem
This is an interesting discussion, as it provides a generalized and broadly applicable framework for why we often see things where they don't indeed exist.
Here is a refutation by one of the leading "hoax" proponents: http://hydeandrugg.wordpress.com/2014/02/06/hoaxing-the-voyn...
In it, he explains the prevalence of half-baked origin theories as a function of several psychological and probabilistic phenomena:
* Pareidolia * Confirmation bias * The birthday problem
This is an interesting discussion, as it provides a generalized and broadly applicable framework for why we often see things where they don't indeed exist.
IIRC, the distribution of letter frequencies doesn't really fit with it being an alphabet for a natural language.
I believe that presumes it's a traditional phonetic alphabet. I'm not a linguist by any measure, but the inventiveness I have seen in human languages gives me hope here.
But let's assume this is an elaborate hoax by a masterful con-artist. Are you really claiming that someone spent so long crafting such a forgery, with nothing before or after - like some black dahlia of literature?
But let's assume this is an elaborate hoax by a masterful con-artist. Are you really claiming that someone spent so long crafting such a forgery, with nothing before or after - like some black dahlia of literature?
> like some black dahlia of literature?
What do you mean by this? I haven't heard the expression before....
What do you mean by this? I haven't heard the expression before....
> Statistical analysis of the text reveals patterns similar to those of natural languages. For instance, the word entropy (about 10 bits per word) is similar to that of English or Latin texts.
http://en.wikipedia.org/wiki/Voynich_manuscript
http://en.wikipedia.org/wiki/Voynich_manuscript
Unfortunately, entropy measures don't reliably distinguish linguistic scripts from other symbol systems:
http://languagelog.ldc.upenn.edu/nll/?p=4652
Wikipedia says that the manuscript has been dated to the early 15th century.
Is carbon dating sufficiently imprecise?
Is carbon dating sufficiently imprecise?
I believe the carbon dating is based on the fabric, which can cause some imprecision (the fabric can be dated as significantly older).
Ah of course!
Thank you :)
Thank you :)
Reading the paper, they make a lot of assertions that are a bit too confident like this plant is most definitely X and this is obviously Y. This is an interesting idea but just because they resemble some plants doesn't make them those plants all things considered.
Original article http://cms.herbalgram.org/herbalgram/issue100/hg100-feat-voy...
Original article http://cms.herbalgram.org/herbalgram/issue100/hg100-feat-voy...
Previously on HN: https://news.ycombinator.com/item?id=7108472
The top comment on this link has a nice explanation of why the offered explanation is unlikely to be a sufficient one. I find the claims intriguing, but a full decryption would definitely take more.
Here's someone else taking a http://www.ciphermysteries.com/2014/01/21/brand-new-new-worl... go at the paper, too.
Well, it's a good theory. Being a manuscript from Mexico, it wouldn't be surprising if it had eluded many researchers, who often have an European-centric background. In Brazil there are also manuscripts in the native's, non-written language (tupi) codified by catholic priests dating back to XVI, but those are better understood because the language was kept alive up to the late XIX. If it weren't that, it would be undecipherable by now, maybe called a hoax too.
"If it was a hoax, then we have to invoke Jules Verne and his time machine."
I wonder if he meant H.G. Wells, or if that was a reference to Back to the Future III.
I wonder if he meant H.G. Wells, or if that was a reference to Back to the Future III.
Either way, I enjoyed the underlying sentiment:
when you're trying to analyze text from documents of uncertain origin, one of the most powerful tools is to look for information that would be known to someone living in a particular place and time, but that wouldn't have been known to "outsiders" (ie, forgers) until after the manuscript was discovered. For example, if you find that a manuscript discovered in the early 1900s uses the right statistical distribution of names from a particular area in the 1600s, and that name distribution was determined from more recent archaeological finds, that suggests the document was written by someone from that time and place.
when you're trying to analyze text from documents of uncertain origin, one of the most powerful tools is to look for information that would be known to someone living in a particular place and time, but that wouldn't have been known to "outsiders" (ie, forgers) until after the manuscript was discovered. For example, if you find that a manuscript discovered in the early 1900s uses the right statistical distribution of names from a particular area in the 1600s, and that name distribution was determined from more recent archaeological finds, that suggests the document was written by someone from that time and place.
That's like a message with a cryptographic signatures carved into the fabric of our planet. Awesome.
Obligatory http://xkcd.com/593/
I was going to write something similar, but nickpelling's comment is better than what I would have come up with:
"The shockingly dull truth is that, if you compare what these two authors have published in an American herbal medicine quarterly with the hundreds of bad Voynich theories out there, they're doing basically the same thing in basically the same way:-
"(1) fixating on a single observation that just happened to excite their imagination
"(2) doggedly following the logic of that single observation to the end
"(3) ignoring or dismissing all evidence that would seem to contradict their story"
"The shockingly dull truth is that, if you compare what these two authors have published in an American herbal medicine quarterly with the hundreds of bad Voynich theories out there, they're doing basically the same thing in basically the same way:-
"(1) fixating on a single observation that just happened to excite their imagination
"(2) doggedly following the logic of that single observation to the end
"(3) ignoring or dismissing all evidence that would seem to contradict their story"
The Voynich manuscript is one of my favorite mysteries. In a lot of ways, I hope it's never really solved.
Lost my interest when he mixed Jules Verne with H G Wells.
We do have a reasonable Nahuatl corpus, and it remains a (very minor) spoken language, so if it is such, there's a good chance it could be understood. But there's plenty of extinct indigenous languages with very little or no attestation, and if it's one of those, well, hopefully it's related to something we still know.
The orthography remains a problem, but it's quite plausible that some friar somewhere in Mexico came up with this European-derived script to transcribe the local language, and it didn't stick for any number of reasons. There are several non-standardized Latinesque orthographies we know (see Carochi) for Nahuatl, so it's a known phenomenon. (And there's still no standard system!) The hand of the Voynich seems reasonably practiced, so it must've had some use, but it could possibly still have been only one man over a few years. (It's clearly not an indigenous script.)
The friar involved could perhaps be called too clever for his own good; imagine if St Cyril wrote a few books in his script and then died of bad gruel, and some Pope got hold of them. Same sort of thing. If the (theoretical Spanish friar) Voynich author had been less inventive, we'd (if this theory is true) have long since understood it.
Anyway, hopefully a few Meso-American language experts can take a look and decide if there's a possible match.