Re: Request For Status: Language vs. Cipher

[Mark Perakh <perakh@xxxxxxxxxxx>]

Rene wrote:

> On Stolfi's post:
>
> > [...] I don't believe it is purely random text (glossolalia, madman's
> > drivel, etc.), because it has too much structure and homogeneity, and
> > I cannot see how those features could be faked (or even perceived) by
> > a 15th century author.
>
> I do not think it's random either. What's also important is that the
> structure depends on the context. An apparent list of star names in the
> Ms has almost unique words (in the sense that they do appear elsewhere
> in the Ms but are almost unique withing the list). The list is 300 items
> long and I do not think there is any other stretch of 300 words in the Ms
> with so little repetition.
>
> > I cannot believe it is a fraud, either, because the feeling is all
> > wrong: it would be like counterfeiting a 3-cent coin. A scholarly
> > hoax, say on Baresch or Kircher, is somewhat more likely; but that too
> > has its problems.
>
> For a scholarly hoax it is *far* too long.
> (But we should always be careful with 'feelings', since our feelings
> belong to a different world than those prevalent in the 15th-17th
> Century.) If it's a fake, it would have to be for monetary gain,
> or for elevated status. In modern terms: like writing a fake operating
> system in order to gain millions. (Sorry, Mike)
>
> > So, almost by exclusion, my current opinion is that Voynichese is a
> > rather straightforward encoding (e.g. a phonetic transcription) of
> > some "exotic" language; and the structures that we see in it are
> > basically those of the language itself.
>
> Let me surprise you by saying that I tend to agree. I would add that
> the structures may also come from the writing system rather than the
> language itself. Of course, my definition of exotic differs a little
> from Stolfi's.
>
> > The "exotic language" theory seems quite plausible historically
> > (indeed Baresch himself apparently believed in it), and seems to
> > explain many features of the VMS that are hard to explain otherwise.
> > For instance, why the text does not include any words in "classical"
> > scripts (Roman, Greek, or Hebrew), why there are no number-like
> > symbols, why we don't see any grammatical structure, why the plants
> > and cosmology look so alien, etc..
>
> Apart from the alienness of plants and cosmology, which is open to
> debate, the above are all valid points in favour of a translation
> (or transcirption or encoding) of an 'exotic' language. Depending
> on the extent of the author's world, exotic could also include
> Hebrew, Arabic, Syriac or Persian, which would not distinguish numerals
> from alphabetical characters, and which should not be expected to
> include Latin or Greek words or symbols.
> There are more exotic possiblities beyond the above, still not being
> quite as exotic as the far East...
>
> > An invented language is also a possibility, of course. However, it
> > seems that invented languages are either utterly logical, and hence
> > utterly unnatural (like Dalgarno's, if I got it right, or
> > Loglan/Lojban); or quite similar to the natural languages known to the
> > inventor (like Hildegarde's, Esperanto, Enochian, Klingon, etc.).
>
> Of these, Hildegarde's is most interesting chronologically. Of course,
> she only invented a long list of nouns. In the 2 centuries leading up
> to the creation of the VMs, someone just might have taken the
> next step...
>
> Let's also not forget Roger Bacon, who claimed to have devised a
> 'common language' using which he thought he could teach people other
> languages in only a fraction of the time needed using a standard
> method....
>
> >     > There are a great many papers on entropy and Zipf's laws, but is
> >     > there also already a summarized (somewhat final) conclusion of
> >     > the findings?
> >
> > Those studies generally show that there is nothing terribly wrong
> > with Voynichese as a natural language, although it doesn't seem
> > to be a standard one.
>
> Let's say that in most statistics, the VMs text scores outside the
> interval occupied by 'normal' languages. (I'm thinking of languages
> written in an alphabetical script with 24-36 characters).
>
> > Some statistics change (sometimes radically) from section to section,
> > while others, including the basic "word" structure, are surprisingly
> > constant.  There is good evidence that the pages and sections were
> > bound in the wrong order.
>
> What seems certain is that they were not bound in the order in which
> they were written. The next step implied above is a very reasonable
> interpretation. (i.e. I also think they are bound in the wrong order.
> I even think we can restore the order, with a little more effort).
>
> >                                                  [...] most people
> > working on decipherment seem to be assuming a complex code. Under that
> > approach, there is not much chance of doing "a little progress" --
> > either you crack the code, or you stay stuck at the beginning.
>
> Yes. And worse: most people who have proposed a solution in the past,
> appear to have at one point assumed a theory including author, language,
> contents of the MS and everything, and then started translating.
> When counter-evidence showed up, this was either discarded or molded
> to fit the theory. (Of course, there may have been those who realised
> they were on the wrong path and we just never heard of them).
>
> I think that what some of the list members are doing, namely analysing
> the structure of the words, the script, and the phrases, is the
> one way in which incremental progress is possible. Little bits of
> evidence (or just strange features) have been found over the last
> decade and maybe, hopefully, one day someone has the idea that puts
> it all together.
>
> >     > Is the text of Newbold's "decipherment" and possibly the
> >     > counter-arguments available online or digitally?
> >
> > I don't know whether the text is available through the net. (In fact I
> > don't know whether he deciphered more than a few sentences.)
>
> The one thing that may be said about Newbold's solution is that he has
> produced a good quantity of grammatical (I think) and totally sensible
> plain text. He even produced plain text which included information
> that he was supposedly not aware of and which could be verified afterwards.
>
> This brings me to a question I've been wondering about for some time,
> and which I will address in a separate post.
>
> Cheers, Rene

In response to the most recent exchange of views, apparently prompted by
Andreas' questions, I take the liberty of reminding the following: A few months
ago I suggested some thoughts about VMs based on the application of Letter
Serial Correlation (LSC) test.  The only response I ever received was from
Rene, whose opinion was that the two articles on my website were relevant to
the discussion.  I don't know if anybody besides Rene has ever read those two
articles (Frogguy once said he would do so, but never said he actually did) and
if anybody did, was the absence of reaction due to the opinion that it all was
crock or that it was just nothing new, or to the unwillingness to delve into
LSC method..  All three views/approaches are legitimate.  However, I would like
to say that the LSC test in itself was proven to reasonably well distinguish
between meaningful texts,  gibberish and artifically created texts with either
low or high entropy.  The application of LSC to VMS resulted in the data
exactly like those obtained for meaningful texts in 12 natural languages. The
LSC tests also revealed a considerable difference between VMS-A and VMS-B,
leading to a conclusion (arguable, of course) that both were written in the
same language but A using much more abbreviations than B.  The empirical
measure of the overall entropy, based mainly on LSC data, placed both A and B
right within the range for those 12 languages.  On the other hand the letter
frequency distribution (assuming every symbol in VMS is a letter) was much more
non-uniform than in any natural texts explored. There was also a division into
vowels and consonants suggested (again, arguable). That is a brief summary of
those data.  Of course I would rather see rebuttals than silence, but
apparently it is not going to happen. Anyway, I had fun doing that study, so it
was not completely in vain.  Best to all.  Mark