3 ms·

One theory is that it's a 15th century hoax. That fits the timeline reasonably well. That strikes me as being more plausible than the idea that it was a later

One theory is that it's a 15th century hoax. That fits the timeline reasonably well. That strikes me as being more plausible than the idea that it was a later hoax. Depending on how much of the provenance you believe, that would place its creation sometime between the early 17th and late 19th centuries. That does stretch credibility a bit. The hoaxer wouldn't just have to find some old calfskins; they would have to find an entire book's worth of unused (I believe it's been confirmed that the book is not a palimpsest) calfskins that all date to around the same time. And then they'd need to get all the other materials right, which is something that, depending on where in the timeline you are, may have required knowledge that was not available at the time. The main problem that I'm aware of with it being a 15th century hoax is that it's hard to imagine why someone would do such a thing. But this was around about the renaissance, where there was a revival in interest in obscure and ancient texts, and rich people who weren't monks were starting to get interested in books. So perhaps there was something of an underground market for these sorts of things, which would at least permit a financial motive.

dane-pgp 6y ago
But doesn't the text have language-like statistical properties that are hard to generate by chance? I suppose that any sufficiently convoluted random/hoax process could result in a text which had the necessary properties, but it seems unlikely that the hoaxer just happened to have picked a process which was able to convince not just their target market but also generations of statisticians who had not yet been born. On the other hand, I understand that some of the statistical properties (at the line or page level?) are quite unlike those that are found in known languages, which might be the fingerprint of a process which doesn't preserve meaning, and therefore suggests a hoax.
- mumblemumble 6y ago
  It's not consistent with someone scribbling at random. (See the Codex Seraphinianus for an example of what that looks like.) But some researchers have demonstrated formulaic ways of generating nonsense text with the same kind of characteristics, using methods that were known in the 16th century. Adding support to that hypothesis is that there is no evidence of any corrections anywhere in the book. If it were real text, you would expect mistakes and corrections. If it were meant to be nonsense, there would also be mistakes, but there would be no particular need to correct them.

dane-pgp 6y ago

But doesn't the text have language-like statistical properties that are hard to generate by chance? I suppose that any sufficiently convoluted random/hoax process could result in a text which had the necessary properties, but it seems unlikely that the hoaxer just happened to have picked a process which was able to convince not just their target market but also generations of statisticians who had not yet been born. On the other hand, I understand that some of the statistical properties (at the line or page level?) are quite unlike those that are found in known languages, which might be the fingerprint of a process which doesn't preserve meaning, and therefore suggests a hoax.

mumblemumble 6y ago
It's not consistent with someone scribbling at random. (See the Codex Seraphinianus for an example of what that looks like.) But some researchers have demonstrated formulaic ways of generating nonsense text with the same kind of characteristics, using methods that were known in the 16th century. Adding support to that hypothesis is that there is no evidence of any corrections anywhere in the book. If it were real text, you would expect mistakes and corrections. If it were meant to be nonsense, there would also be mistakes, but there would be no particular need to correct them.