25-08-2022, 10:34 PM
There's also the You are not allowed to view links. Register or Login to view. of the presentation - also in Russian.
I watched it briefly, the idea of the author, if I got him right, is that vowelless rank vs frequency letter distributions tend to cluster for languages from same language groups. E.g. distributions for German and English (both belonging to the Germanic group) are closer to each other than e.g. English and French.
Ultimately he concludes that the VMS seems to be bilingual (some folios one language, some in another) with Danish and Latin being the best candidates (selected from European languages). There's a slide (Figure 7 in the paper) plotting the supposed languages against the folio space - it would be interesting to check this supposition versus Currier A/B, but apparently the author does not do that.
He also employs other statistical methods as indicated by RobGea (Hurst and spectral portraits) shown to support the idea of bi- (or multi-) lingualism of the VMS text.
It is true that he does not touch the entropy issue, at least in the video.
There's been a previous publication by the same author, and for sure it was discussed here in the News section. I had a recollection that back then I had been sceptical about it for some reason, but could not recall the exact one. Upon the glance on this new paper, however, I instantly recollected what the matter was - he still claims that the VMS was produced in 16th century. It is a bad sign, in my opinion, for any serious investigation of the VMS to be mistaken in such very basic facts about it.
I watched it briefly, the idea of the author, if I got him right, is that vowelless rank vs frequency letter distributions tend to cluster for languages from same language groups. E.g. distributions for German and English (both belonging to the Germanic group) are closer to each other than e.g. English and French.
Ultimately he concludes that the VMS seems to be bilingual (some folios one language, some in another) with Danish and Latin being the best candidates (selected from European languages). There's a slide (Figure 7 in the paper) plotting the supposed languages against the folio space - it would be interesting to check this supposition versus Currier A/B, but apparently the author does not do that.
He also employs other statistical methods as indicated by RobGea (Hurst and spectral portraits) shown to support the idea of bi- (or multi-) lingualism of the VMS text.
It is true that he does not touch the entropy issue, at least in the video.
There's been a previous publication by the same author, and for sure it was discussed here in the News section. I had a recollection that back then I had been sceptical about it for some reason, but could not recall the exact one. Upon the glance on this new paper, however, I instantly recollected what the matter was - he still claims that the VMS was produced in 16th century. It is a bad sign, in my opinion, for any serious investigation of the VMS to be mistaken in such very basic facts about it.
![[+]](https://www.voynich.ninja/images/collapse_collapsed.png)
