Crossing the Line from Amateur to Dilettante

Although in general I recommend the Voynich mailing list to everybody interested in the subject, the current discussion about using distributed computing to crack the VM with a statistical brute-force attack very much reminds me of people planning some open-heart surgery with their medical knowledge gained from watching a few episodes of House.

“If at first you don’t succeed, use a bigger hammer.” The sad thing is that while one supposed side effect of this would be to make the project and VM research more public (in the vein of the SETI project), and hence to attract more people to it, I’m afraid a bungling approach like what is currently considered would have the opposite effect of exposing us to ridicule.

(Not to mention the fact that people seem to think that using a distributed approach would invalidate all previously gathered statistical information about the VM, like that it certainly is not a simple substitution, and that with a high degree of confidence one VM word is not equivalent to one plaintext word.)

Advertisements

15 thoughts on “Crossing the Line from Amateur to Dilettante

  1. Hahaha! More like trying to climb Everest using tampons for crampons. :^p

    As you say, Voynichese is certainly not a simple substitution cipher, though I’d finesse your second assertion to “with a high degree of confidence one VM word is not a letter-for-letter equivalent version of one plaintext word.”

    I still don’t grasp why people get so upset at my suggestion that Voynichese is essentially a verbosely enciphered shorthand. It’s not as if it requires techniques unknown circa 1424, eh?

    PS: have you ever wondered why ‘abbreviation’ is such a long word? I know I have. :-)

    PPS: I’m halfway through writing up a post on your personal favourite stroke theories (if my website ever comes back to life, that is *sigh*)

  2. PS: I’d say that SETI has a higher chance of decoding an alien recipe for eating humans medium-rare in bearnaise sauce than BOINC has of deciphering the VMs. :-)

  3. Well, to be honest, I don’t see this project really moving in a direction as if it would soon (read: ever) take off. At the same time, it is clear that not everybody understands what is being proposed by the various people. There are actually valuable things that could be done, but since I am not in a position to be actively involved in that, I will hold back my vocal horses :)

  4. Nick, on the odds of SETI vs. Voynich decoding (your second comment):

    I’d say yes and no.

    Yes, because the SETI people actually have a system running and the Voynich community doesn’t. It’s as simple as that. I don’t know if it is more weird to search for Alien life or to decode the Voynich MS, but the former group clearly has had the critical mass needed to include enough driven people to set up such a scheme.

    No (and this is clearly just my opinion), because the probability that there is actually anything to be found at all, is diminishingly smaller in case of SETI compared to decoding the Voynich MS.
    And underestimating others is usually a grave mistake…

  5. @ Nick: I’m waiting with bated breath for your thoughts on the strokes, but access to the Ciphermysteries site has been intermittent at best lately…

  6. Elmar: the hosting company I use is having a bit of a bad time at the moment, not sure how long this will continue.

    Just so you know, once your WordPress blog database gets past a certain size (20MB? 30MB?), backing it up can become quite problematic: hence, doing something about a failing host isn’t as easy as you might think. :-(

  7. Elmar, from what I understand of your stroke theory, I would say that, if the Voynich MS was encoded in this way, the algorithm I had in mind for a distributed attack on the text would have a chance of finding it.

    Proably not in all cases though.

    Again from my limited understanding of Philip Neal’s method, this would also have a chance of being cracked by the same method.

    Furthermore, if the Voyinch text just has lots of nulls, it would also have a chance of being cracked…

  8. No amount of cryptanalysis will decipher the VMs if a wrong-headed set of assumptions is used to inform the choice of what kind of thing to search for.

    For example: is the wide & shallow Voynichese dictionary an artefact of polyalpha-like statefulness (as Steve Ekwall [of all people!] suggested, though probably not as precisely), or is it the result of conscious contrariness on the part of the encipherer (i.e. not using the same words twice, or continuously abbreviating words differently)?

    Put like that, perhaps a sensible use of a staggeringly large amount of computer resources would be to try to answer the question “is Voynichese stateful?”

    Which is to say, “can we identify any set of Voynichese letters which seem to modify an internal state machine?” Of course, the gallows characters are a natural first choice, but they are by no means the only choice. For example, I suspect that the “o + gallows” (but not “4o + gallows”) and “y + gallows” pairs encipher different tokens entirely, which would probably be enough to confuse any statefulness analysis run.

    Note that I’m not saying that a machine would be able to decipher the VMs using this: rather, that if we can use brute force to identify the statistical fingerprint of some kind of statefulness, then we stand a vastly better chance of working out what it is.

    I would add the caveat that I suspect Currier A paragraphs, Currier B paragraphs, and labels operate according to slightly different rules, and so test runs should probably be done on a label-less A corpus and a label-less B separately.

    Cheers, ….Nick….

  9. René — I think the Strokes theory should be testable with a computer attack (after all, it’s essentially just 2*26 snippets, give or take a few, which should make up the text), while I don’t think it warrants a distributed attack with massive force. Any decent PC should do, with a good algorithm behind it (which is still a problem for me) ;-)

    I agree that *any* particular theory should be testable by a cloud computing attack, but the attitude on the list apparently was that even *without* pursuing one particular idea, it would work out, ie the computers would spit out the correct answer even without anyone knowing the correct question.

    Which I doubt.

  10. Hi Elmar:

    “…but the attitude on the list apparently was that even *without* pursuing one particular idea, it would work out, ie the computers would spit out the correct answer even without anyone knowing the correct question.”

    It may be that this assumption was in there somewhere, but I think the general tone of that discussion was in the form of a question, with several proposed answers. I think one conclusion, if there were any, was that it would be necessary to know “the correct question”… i.e., reasonable algorithms which would give a chance of success, by assuming a structure or set of structures… before either applying this attempt, or knowing if it would be feasible, or knowing if mass computing would be necessary.

    Said simply (painful for me): The problem was being asked and discussed, with some ideas, but no firm conclusions as to whether it would work, or exactly how to proceed. I think there were many rational ideas in that discussion. I saw it as very informative, even though not decisive (yet). Rich.

  11. Hello Elmar,

    you raise an important point. Nowadays, there’s a lot that one can do on a modern PC and in 24 hours, and a distributed approach would not be necessary in many cases.
    It’s really only the right approach for problems which requiring searching in a multi-dimensional solution space, or where every single ‘test’ takes more than just a fraction of a second.

    Of course, one should see this type of attack not as an attempt to solve the Voynich MS, but to test specific ideas (which can of course also be rather general).

  12. Dobry den. Na webu jsem našel( obrazek ,Evidence 3:Castillum okoe 89). Na obrazku je hrad a okolo oblacky. A 2 slova. Není to okoe 89. Vyznam tohoto obrazku = Varovani. (Nebezpeci). 2 slova po prelozeni.(rukopis = oMccdq oMaedq ).Preklad = z mlg ti da stin.(mlg = mlha,vypary,dym, Kazdá kyselina vydava dým,kouř,mlhu,mlg. Oblacky okolo hradu jsou výpary. Autor poukazuje na nebezpečí při práci s kyselinou. Zde se jedná o kyselinu dusičnou (oM = ocot M = ocot= kyselina, M = D = Dusičná, ( M=D,cabalistický system).Jelikož autor použil gematrii tak 2 slovo také znamená, (rukopis = oMaedq= oM jedy = ocot M jedy = kyselina Dusičná jed.Autor těmito dvěma slovy říká = při práci s kyselinou se můžeš přiotrávit a zemřít.( zemřít= stín). Teď už tomu možná porozumíte. Na těchto 2 slovech je vidět,jak je důležitá pro překlad čeština.

  13. Pro překlad je důležitý czech jazyk (čeština)! Né anglický jazyk! Špatně to překládáš. Jak by jsi chtěl dešifrovat manuscript, když nepřeložíš ani moje slova.A jsou to slova současná, ne stará 400 let. Teď ještě k tomu obrázku. Na obrázku je namalován hrad.(všechny obrázky v manuscriptu jsou symbolické !!)Takže hrad = athanor!!! (Athanor = alchymistická pec). Těchto pecí bylo v alchymistické laboratoři několik, většinou 3. Samotný tvar těchto athanoru (pecí) silně připomínal věž hradu. Další velice důležitá věc pro dešifrování. Slovo 3 znaky manuscript.Znaky = d,a,M. ( dám). Vy to špatně překládáte. Není to přehrada,ale význam slova v czech jazyce = vložit. (vložit do alchymistické nádoby)!! Hlavně žádnou prehradu. Co by asi dělal alchymista s přehradou ??!! Komunikace přes tyto stránky silně připomíná komunikaci s N.Pellingem.Ten umí dešifrovat velký kulový.(Hodně velký kulový) Zlatoděj J.T. Czech republic

  14. Zdravim všechny. Přeložím jakoukoliv stranu manuscriptu. Pánové když dáte na web dobře naskenovanou stranu ( COSMOLOGICAL ) ,tak vám ji do detailu přeložím. Vše co jsem našel na webu ohledně této strany je špatně nafocené a tak to nelze přeložit.Zlatoděj J.T. česká republika Plzeň.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s