There are various models for the structure and spelling of Voynich vords. None of them are entirely successful in describing every permutation we find in the text. In every model we find non-conforming vords. But the fact that any such model is even moderately successful is itself significant and says much about the nature of the text.
The model below is not my work. The hard work was done by the insightful Thomas Coon who devised his model - the Voynich Vord Verifier - to try to explain in a simple table the various perplexing features visible in the text. I have modified Mr Coon's work, adding some new features, giving a different emphasis to it, comparing it to and combining it with some other models.
I point out that, like many other models, it fails at the first hurdle: the first vord in the manuscript fachys is non-conforming. The paradigm has its limitations. Otherwise, it maps enough Voynich vords to be useful as a study tool. I am not claiming that the author used this arrangement to generate the text. It is a device used to explore aspects of the text. It tries to approximate an implicit and pervasive pattern that is characteristic of Voynichese. The places where it fails are perhaps more interesting than those where it succeeds.
Every column can be passed over as a null, as can each group A, B, C or Final.
The model does not account for many of the constraints under which various glyphs are bound in certain circumstances. As well as the table, you need a set of rules.
The glyphs in red are GALLOWS or PEDESTALS which are marked as a separate class of glyph. They are a special class and are at the core of the paradigm.
The heavy line indicates the default VORD BREAK.
The default vord is QOKEEDY which repeats because q prefers to be followed by o, o by k, k by ee, ee by d, d by y, and y by q with a vord break.
The natural division of the glyphs is into consonants and vowels and groups of CV syllables, taking o e y and a as vowels (with CVV and CCV possible.)
There are three sections and a group of FINALS. These are an extension of group C which can otherwise produce final glyphs. The Finals may have the form vowel-consonant (VC). Group C can combine with the Finals.
The glyphs based upon the minim or backslash [\] = i, ii, iii etc. are confined to group C and the Finals. They come at the end of the sequence. [r] and [l] can be exceptions.
We discover that the [l] glyph in group A is problematic.
Vords can be made from groups A+B+C, A+B, B+C or A+C.
Where a vord can be made several ways, choose the most compact method with intact CV syllable groups.
The vertical columns have no organizing principle at this stage, except the gallows which are ordered by frequency top to bottom.
Making as many vords as possible is not the objective. The model is adapted to exploring patterns that accompany the QOKEEDY cycle and the observation that QOKEEDY is consonant/vowel/consonant/vowel/consonant/vowel. This suggests consonant/vowel as a basic dichotomy.
The model does not distinguish between Dialect A and Dialect B.
Doubtless the whole thing can be refined further.
* * *
The test of any model is to see whether it conforms to an observable textual phenomenon or not. For example:
It is an observable and very solid rule of the text that there must be a VORD BREAK after y, except where y appears before t or k.
Similarly, we can take any rule, law or observation and test the model against it.
Consider a more complex proposal such as Emma May Smith’s account of “y deletion.” See her work here.
In certain environments – specifically in the middle of vords – y is deleted or dropped. Instead, she argues, a and y are interchangeable.
We can see this in our model. The vowel set of group A is: e, o and y. The vowel set of group B, however, is: e, o and a. y is excluded from group B and a appears instead. The model depicts “y deletion.”
* * *
SOME EXAMPLES OF PARSING BY THIS MODEL
qokaiin = qo - ka - iin
okeody = o - keo - dy
qokchdy = qo - kch - dy
chokor = cho - ko - r
qokodaiin = qo - ko- da - iin
shodaiin = sho - da - iin
kchdal = kch - da - l
chopo = cho - po
qoeees = qo - eee - s
ysheey = y - shee - y
chocpheod = cho - cpheo - d
checkhdy = che - ckh - dy
lkeeedy
= l - keee - dy
sheedo = shee - do
daiin = da - iin
ofchedy = o - fche - dy
chokchol = cho - kch - ol
pchodaiin = pcho - da - iin
We can use the model to look closer at non-conforming vords and establish the elements that do not conform. For instance:
psheot
We parse it as:
pshe – o – t
The problem is only the final [t]. l r m and g were possible, but not the gallows t. Otherwise it conforms.
Another example:
ykolairol= y- ko – la – [i]ro – l
Here the combination –iro- is non-conforming, although it is only the [i] that cannot be accommodated. The vord fails to conform by one glyph, [i], indeed one stroke. We find that many non-conforming vords deviate from the model only slightly.
R. B.
No comments:
Post a Comment