Access to the paper makes following this conversation much easier: many thanks to Torsten for sharing it!
It seems that much of the exchanges have focussed on Figure 4. Timm and Schinner present two Voynich samples and an English sample:
You are not allowed to view links. Register or Login to view. Nablator has computed a similar graph for a text by Aristotle translated into Latin by Boethius.
If I understand correctly, when line distance increases from 0 to 60:
Isn't the curve for the Latin Aristotle much more similar to that for the VMS Recipes than to that for Alice in Wonderland?
As Nablator points out, the text by Boetius is peculiar; yet I don't think it can feature the consecutive repetition of identical words (which is quite rare in Latin prose). Clearly, exact reduplication must contribute to the effect discussed by Timm and Schinner (even if of course other phenomena contribute as well). It would be interesting to see how texts in a reduplicating language behave, but, in most examples I have seen, reduplicated words are joined into a single word: this of course would make them invisible to this method.
It seems that much of the exchanges have focussed on Figure 4. Timm and Schinner present two Voynich samples and an English sample:
Quote:Black line (a): VMS “Recipes” section (f103r–f116v, Currier B);
red line (b): VMS “Herbal” section (f1r–f66v, Currier A);
blue line ( c ): the first 10000 tokens of “Alice in Wonderland”;
dashed lines: the respective asymptotic values of M̄ .
You are not allowed to view links. Register or Login to view. Nablator has computed a similar graph for a text by Aristotle translated into Latin by Boethius.
If I understand correctly, when line distance increases from 0 to 60:
- In the VMS Recipes, word-distance increases by 0.20 ca (from 4.75 to 4.95)
- In the VMS Herbal A, word-distance increases by 0.28 ca (from 4.4 to 4.68)
- In Alice, word-distance is nearly unaffected (it could increase by 0.01 ca)
- In the Latin Aristotle, word-distance increases by 0.17 ca (from 5.4 to 5.57)
Isn't the curve for the Latin Aristotle much more similar to that for the VMS Recipes than to that for Alice in Wonderland?
As Nablator points out, the text by Boetius is peculiar; yet I don't think it can feature the consecutive repetition of identical words (which is quite rare in Latin prose). Clearly, exact reduplication must contribute to the effect discussed by Timm and Schinner (even if of course other phenomena contribute as well). It would be interesting to see how texts in a reduplicating language behave, but, in most examples I have seen, reduplicated words are joined into a single word: this of course would make them invisible to this method.