11-10-2019, 08:47 AM
(10-10-2019, 11:05 PM)Stephen Carlson Wrote: You are not allowed to view links. Register or Login to view.I'm wondering if Voynichese was meant to be pronounced or pronounceable.
If it was, it would fit an unknown-language or glossolalia hypothesis, but not so much a ciphertext hypothesis. The answer seems neutral on a hoax hypothesis.
In favor, the glyphs really do look like an alphabet (some glyphs even present in Latin MSS), the "vords" seem to have an onset-nucleus-coda structure (which is pronounceable), and the EVA representation is pronounceable.
Thank you, Stephen! Your question points to a specific problem that is of the greatest interest for all Voynichese theories.
Your argument about EVA seems quite strong to me: the script can be mapped into sounds in such a way that the result is pronounceable. It either is so by-design or coincidentally. A coincidence seems unlikely for something as complex as a phonetic system. Of course, EVA was designed to be pronounceable, but, for a random text, it would have been impossible to achieve good results.
A while ago, I compared the performance of a few written texts in terms of the level of alternation that can be reached by You are not allowed to view links. Register or Login to view. into two classes that basically correspond to vowels and consonants. This was inspired by You are not allowed to view links. Register or Login to view. on the application of Sukhotin's algorithm to Voynichese. What I found is that levels of alternation in the VMS are as high or higher than the linguistic texts I examined. Results of course depend on the transcription system: the better results require treating repeating EVA characters (i-sequences, e-sequences) and benches as corresponding to a single sound.
The You are not allowed to view links. Register or Login to view. Emma and I wrote also points in the direction of a pronounceable text: the same kind of alternation that is observed within words also occurs across spaces between adjacent words. For instance, the dislike for identical glyphs to repeat inside words can also be observed across words: oo and dd are rare inside words and word sequences like sho.okeeo and ched.daiin (where '.' represents a space) also are rarer than expected.
(10-10-2019, 11:05 PM)Stephen Carlson Wrote: You are not allowed to view links. Register or Login to view.Against, no phonetic values are known,
It seems to me that the state of things you describe is necessary for this discussion to be meaningful. Per-se, it does not disfavour the idea of glossolalia or an unknown phonetic language.
(10-10-2019, 11:05 PM)Stephen Carlson Wrote: You are not allowed to view links. Register or Login to view.the script has layout effects (e.g., first character of lines, first lines of paragraphs, etc.)
I agree that there are several features that seem difficult to reconcile with a linguistic interpretation, and line-effects are among these. Line-effects can be observed (for instance) in the case of poetry, the usage of abbreviations (which tend to be more frequent at line-end), embellishments like You are not allowed to view links. Register or Login to view. (which are more frequent in the first lines of paragraphs) and hyphenation (which results in pseudo-words at line end/start which do not appear anywhere else). Some of these phenomena might concur to line-effects in the VMS, but I am not sure and I don't think they can explain everything. On the other hand, we know little of glossolalia. If one was producing glossolalia by speaking (in his head or aloud) while writing, there could be other unintentional effects at line boundaries. Sadly, this speculation is difficult to proof or disproof. Transcriptions of glossolalia could allow us to understand more of things like the level of word reduplication, but would not help with line-boundaries effects.
About JKP's observations with respect to the positional rigidity of characters in Voynichese, it is important to remember the difference between script and language. He writes that "letters in natural languages move around within words. For example, the letter "a" can be in any position in a word", but he knows very well that in medieval scripts this is not always the case. E.g. the Latin "9" abbreviation typically occurs at the beginning and end of words, short "s" and long "s" are positionally constrained etc. The problem he describes exists, but part of the phenomenon can be explained as a property of the script, rather than the language (see for instance You are not allowed to view links. Register or Login to view. that a and y are phonetically equivalent positional variants of a single sound).
(10-10-2019, 11:05 PM)Stephen Carlson Wrote: You are not allowed to view links. Register or Login to view.many vords seem differ by only a stroke or two and may have been generated by purely graphic means (see Timm & Schinner).
I believe these differences mostly depend on e-sequences and i-sequences. If each sequence corresponds to a single sound, and these sounds are related, differences like these seem to me compatible with glossolalia (or, more optimistically, the degree of spelling variation in a medieval manuscript). For instance an ain aiin aiiin might correspond to different accents, or lengths, of final 'a' (an unknown vowel) - I believe that something similar was proposed by Emma.
Another element contributing to similar (often consecutive) words is the presence / absence of q (e.g. okeey.qokeey). While linguistically this is quite puzzling, I see no problem in believing that similar patterns occur in glossolalia.
The possibly glossolalic You are not allowed to view links. Register or Login to view. includes word sequences like "ard ardo arga arges argah" "umas umascala um" "arcasa arcasan arcusma". I think they are rarer than in Voynichese, and consecutive words appear to differ more than in Voynichese, but they do suggest that quasi-reduplication can be a feature of glossolalia.