[Talk] Upcoming VMS lecture in NYC

[nablator]

Linda and nablator, I agree that this statement is problematic, but not necessarily for the same reasons you do. I think it's just poorly worded. "Secret writing" is an imprecise term that begs for clarification. "A phonetic transcription of sounds instead of letters" comes off as semantic wheel-spinning that doesn't say anything new or meaningful. All letters in phonographic writing systems (alphabets, abjads, and syllabaries) are phonetic transcriptions of sounds, after all.

I have a hunch that what Ms. Davis is trying to say, is that the VMs' text is the result of a scribe simply echoing a speaker's speech in written form, with little to no consideration for what the speaker might have meant to say. In other words, the scribe was merely a scribe, not an editor or an interpreter. Think of a court stenographer or a medical interpreter, both of whom aim to simply echo what another person said, with as little of their effort evident in the finished product as possible. In the setting of a medieval manuscript, one possible scenario which would fit this model is that of a scribe attempting to record a spoken language (s)he didn't understand, especially a spoken language that had no tradition of being written.

It is certainly possible that all the variant spellings that we see in 15th century manuscripts make it difficult to recognize a language, even more so if it is a "dying unwritten language". I have nothing against the idea, it is possible. The problem is that it still contradicts the under-studied facts that very few people seem to notice: that there are many variations in statistical features in the VMS that make it very unlikely to be a "normal" unprocessed language, be it phonetically transcribed or not. Other than the huge gap between the two Currier "languages" there are less visible discrepancies at several levels: quire (for example EVA-ody endings almost disappear in Q13 when they are common everywhere else), page (for example, very common patterns totally missing or overused on some pages) and even sometimes paragraph (for example what ReneZ calls "type-3" and "type-4" words). Of course some statistical variations should be expected especially with small samples, but the variations obviously exceed what is seen in normal texts, where words are not selected for their spelling but for their meaning. A possible analogy would be the type of poetry that uses a lot of assonance and alliteration but the VMS goes far beyond that in scope and extent. The autocopy hypothesis could explain the local self-similarity, but a general drift (that ReneZ noticed in glyph bigram statistics and Torsten Timm and MarcoP saw in common words statistics) needs to be added to have the complete picture.

About the word structure problem that Linda mentions, I also don't understand how it could be solved by the hypothesis of a phonetically transcribed language.