RE: On the Voynich manuscript
Alin_J > 27-07-2020, 09:28 AM
I have read through the whole article once. First of all, there is no mention of either the sources, names or authors of the 'Major European language' texts that were used in the comparison. (Okay sorry my mistake, saw now that they are in the reference list. But they should have put a cross-reference to it). Second, I would like to see also the connected bigram-frequency distribution for the European languages (Fig. 11b), most preferably plotted in the same graph, since the similarity of these distributions is to me the most statistically interesting finding in the whole paper. The unconnected distributions for these texts are shown in Fig. 11a.
There are some other similarly suspicious methodology issues in the handling of the Bigram-Polarity graphs - among the European languages, only the one for German is considered for comparison? Also, is there a stronger rationale for leaving out the -1 to 0 range in the graphs than the somewhat obscurely put "symmetry reason"? And also, the graphs for the VMS herbal section, LL tables and German texts all actually look very similar so I fail to see the point of this whole analysis section with regard to the aim of the study. The fact that the Soyga Tables graph is different is interesting, but beside the point.
Lastly, only the VMS herbal section is analysed. What about the rest of the VMS?
These points to me smell very much like cherry-picking among data/evidence to fit a predetermined theory.
I cannot say anything about the graphology sections, since these are too far outside my area.