[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: VMs: voynich dice game ... sunday thoughts
On Wed, 15 Sep 2004, Jorge Stolfi wrote:
> Here are three arguments that the Voynichese "word spaces"
> are indeed word separators:
> (1) Almost all glyphs are separated from their neighbors by a gap.
> So we are not talking about connected vs unconnected characters as
> in Arabic, but between wider and narrower spaces. While some of
> these width variations may be just optical artifacts due to the
> shape of the adjoining letters, on the whole the distinction between
> narrow and wide spaces is quite marked and seems obviously
> intentional. So, assuming that the narrow spaces are character
> separators, what are the wide spaces if not word separators?
Ths certainly does suggest two levels of significance, and inter-character
(or inter-some characters) vs. inter-word is certainly the most likely
hypothesis to me, though perhaps not the only one. Taken with point
three, however, it seems to eliminate most of the other obvious
possibilities, like constituent-of-syllable-representation vs. syllable,
and so on.
> (2) Since line breaks do not occur at random, it seems reasonable to
> assume that they occur mostly at word boundaries. The statistical
> context of the "word spaces" (i.e. the probability distribution of
> the two adjacent characters) is fairly similar to that of line
> breaks -- provided we allow for the fact that certain charaters like
> EVA "m" (abbreviations?) have a marked preference for the end of
> lines. Therefore "word spaces" are also mostly at word boundaries,
I agree with this, point, too. Note that many languages have verbs in
final position, and in inflectional languages it can be the case that the
inflectional endings of verbs favor a somewhat different set of final
sounds than the inflectional endings of nouns and adjectives (which are
often inflected similarly). For example, -t is, I think, much more likely
to be be verb-final than noun-final in Latin. I think the same - or rather
the reverse - is true of -o, i.e., it is more likely to be verb final.
However, I do not have a cryptographer's appreciation of the statistics of
> (3) The structure and statistics of the "labels" are fairly similar
> to those of the text "words", and most labels contain no "word
> spaces". It seem therefore reasonable to assume that those labels
> are single words of the language, and that the same is true of the
> text "words".
> Note that (3) argues against the "word spaces" having been inserted
> by context-sensitive rules, just to fool code-breakers.
I agree with all these observations.
> Of course the words could actually be numbers in some Roman-like
> number system.
Note that classical numeration schemes are not all Roman-like. There's a
nice description of Greek systems in B.F. Cook. 1998. Reading the Past:
Greek Inscriptions, part of a U of California Press/British Museum series
on ancient writing systems and inscriptions.
The Greek acrophonic system is rather like the Roman one, but using
letters like D = deka for ten, and so on. The alphabetic systgem, uses
alpha, beta, gamma, ..., theta as 1-9, iota, kappa, lambda, ..., omicron
as 10-90, and rho, sigma, tau, ..., vau as 100-900. Modern printers have
schemes for indicating when something is a number, and whether to
understand it as above or below 1000, using accents above and below the
line, etc. The alphabetic scheme dates from the second century BC and was
used into the Byzantine period.
Note that if numerals could be identified, and if they were using an
alphabetic scheme, this might potentially be a very useful key to an
alphabetic orthography, because it would impose an ordering on the
constituent elements of it.
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying: