+- The Voynich Ninja (https://www.voynich.ninja)
+-- Forum: Voynich Research (https://www.voynich.ninja/forum-27.html)
+--- Forum: Voynich Talk (https://www.voynich.ninja/forum-6.html)
+--- Thread: [split] On the validation of Theories (/thread-2886.html)
RE: Unconventional Methods in Voynich Manuscript Analysis - DONJCH - 13-08-2019
(12-08-2019, 04:49 PM)Torsten Wrote: You are not allowed to view links. Register or Login to view.Usually a theory gets validated. For instance it is necessary that a theory is consistent in itself and with the known facts. A theory must also be falsifiable. With other words there must be statements that can disprove the theory. See the following link for more validation criteria: You are not allowed to view links. Register or Login to view.
It might be interesting that this year a course at the university of cologne was given. In this course several Voynich theories were validated (see You are not allowed to view links. Register or Login to view.):
- You are not allowed to view links. Register or Login to view. (Gibbs, Bax, Montemurro, Cheshire)
- You are not allowed to view links. Register or Login to view. (Hauer/Kondrack, Hermes)
- You are not allowed to view links. Register or Login to view. (Rugg, Timm)
Note: I would suggest to split this thread and to start a new thread about the validation of theories.
If there is one theory that stands out as being most likely, I reckon it's probably the Torsten Timm autocopy theory.
On another note, Reference 2, so much space devoted to the Cheshire theory?
RE: Unconventional Methods in Voynich Manuscript Analysis - nickpelling - 13-08-2019
(10-08-2019, 08:33 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view."Proving" is one of these difficult concepts.
I think that it is next to impossible to disprove any of the existing Voynich theories.
The best one can hope to do is to:
- show they are extremely unlikely
- show that they are not based on any evidence
- show that the logic is faulty
(and maybe a few more)
From my point of view, I think it is possible to disprove most existing Voynich theories - for example, I think it is possible to disprove more or less all "pure" linguistic theories, because the way Voynichese works (and indeed, the way the various Voynicheses work) is nothing like English, French, German, Latin, Hebrew, Nahuatl, etc.
The problem comes when people proposing these linguistic theories then go away and hybridize/adapt them to overcome the basic oibjections (e.g. how does language theory X help us make sense of Currier A / B etc? For if it does not, something is badly wrong). These funky hybrids quickly become as complicated than any natural language ever encountered: and the problem then comes that it takes an inordinate amount of time and effort to understand WTF they are (or have become), never mind disproving them as well.
Theories that claim to read the text in some kind of interpretative free-form style yet somehow yielding a readable decryption can normally be disproved by their wonky historical assertions. (Cheshire's text springs quickly to mind here). So there is hope as far as disproving nutty Voynich theories goes.
Cheers, Nick
RE: [split] On the validation of Theories - ReneZ - 13-08-2019
The difference is probably too subtle to make a difference.
I consider the Cheshire thesis to be disproven, and equally most other translations.
Problem is that there always remain several people who don't believe the proof, and others who don't understand it. (Note that Cheshire has followers who believe he is right).
Many of the provenance theories are untestable so they can neither be proven nor disproven.
Exceptions are those cases where they violate either the 15th Century provenance or the presence of the book in Prague for at least half a century. Or in fact both.
These I think one can consider to be disproven.
But is it proof if many people don't believe it?
RE: [split] On the validation of Theories - davidjackson - 13-08-2019
You're never going to please everyone. There are still people out there who believe Elvis is alive, that the Earth is flat and the moon landings were faked.
I say, let them be and don't worry about trying to convince them. There are better things to do with our time...
RE: [split] On the validation of Theories - ReneZ - 14-08-2019
My words exactly David.
RE: [split] On the validation of Theories - Torsten - 14-08-2019
Is there any hypothesis or theory out there that pleases you? If there is one, how did you validate it?
What is with the following hypotheses:
- it is an unknown natural language,
- it is an artificial language,
- it is a shorthand verbose cipher,
- it is a codebook cipher,
- it is an ambiguous cipher,
- it was tampered beyond recognition during copying
Are some of this hypotheses more likely than others? Does one of this hypotheses help us to understand what we already know about the Voynich manuscript? Does one of this hypotheses makes some testable predictions?
RE: [split] On the validation of Theories - ReneZ - 14-08-2019
Torsten,
I really don't understand your question, in the frame of this thread.
It is not about what 'pleases me' (I guess that this is just an unfortunate expression), but about what is provable and/or testable.
Generic classes of solutions are not provable or testable.
RE: [split] On the validation of Theories - Torsten - 14-08-2019
René, "pleases you" was a reference to the previous post of You are not allowed to view links. Register or Login to view. and the "you" is addressing any reader of this thread.
The point of my post was that we can only reject the idea that Elvis is alive, that the Earth is flat and that the moon landings were faked since we know otherwise. As long as the Voynich manuscript counts as unsolved we obviously don't know otherwise. Therefore I was suggesting to look for hypotheses and/or theories that might help us to understand the Voynich manuscript. The hypotheses given in my post are just some examples. If you would suggest some other hypotheses or theories please add them.
Hypotheses about generic classes of solutions are indeed not provable but they should be testable. Otherwise, a hypothesis would not explain anything and would therefore be irrelevant. For instance, a text using natural language should allow us to detect features all natural languages have in common. An encoded text using a verbose cipher or a codebook cipher should result in statistics typical for such a cipher. A text tampered beyond recognition during copying from an external source should not fit too well in the given space or should contain numerous corrections. Only for the inaccurate encoded cipher I would agree that this hypotheses didn't result in any testable predictions.
RE: [split] On the validation of Theories - davidjackson - 15-08-2019
A codebook cipher that was sufficiently large, with an arbitrarily large number of nulls, would not be detectable. It would also be a null solution as there would be no way to decrypt it. We would thus be unable to prove, or disprove, it. And since no solution would appear we would end up discussing it ad nauseum.
It's this sort of irreducible theory that leads to constant pointless argument, which was my original point.
If we take the position that the Voynich is not intelligible, then it doesn't really matter which unprovable route we take to this position - autocopying, lost cipher, gibberish. They are all paths that lead to the same conclusion, that the Voynich has no meaning that we will ever decipher.
But if we take the counter-position (that the Voynich must have meaning, despite all evidence to the contrary) then automatically we reject all of the above theories and must seek to dismiss them out of hand,
Therefore, the real argument isn't over which route to take, as the variables are so large there will always be a counter-position to take.
The real argument is over whether the Voynich has meaning or not. So far, it doesn't and there is no indication that it ever will have.
But again, since this is unprovable (simply a possibility that grows more likely by the year) the circle continues....
Torstein, I never dismissed your theory. I think it's highly interesting, and have said so before. I don't have a particular theory to push, and if you think that I'm comparing your theory to moon landing hoaxes you're very wrong.
RE: [split] On the validation of Theories - Torsten - 15-08-2019
David, you didn't understand my point. I just say that we should validate hypotheses for the Voynich manuscript by looking if they can describe and explain observed phenomena and if they lead to empirically testable and falsifiable predictions. With other words we should handle it like any other hypotheses.
For instance the Voynich faked it hypotheses doesn't make any falsifiable predictions and doesn't help us to explain any observed phenomena which are not explainable otherwise. Therefore such a hypotheses is just irrelevant.
That you reject all hypotheses and theories since someone will always take the counter-position is exactly the opposite of what you said in your last post. A position that all hypotheses are alike since no progress is ever possible also doesn't lead to any falsifiable predictions and is therefore irrelevant.
The Voynich manuscript is full of patterns and the statistical analysis reveals almost perfect results. This means the variables are not at all large, on the contrary it means that the Voynich manuscript is almost perfect example of its kind and that the patterns are very well determined. It is simply a mistake to assume that since some people didn't understand a pattern it must be impossible to understand this pattern. In the same way it is also not necessary that the pattern must contain numerous variables.