10-01-2017, 02:21 PM
I see you make the same point in your latest paper, Torsten, with "the" also generating words like "phe" "khe" and "theee" if the VMS were like English.
The shaded percentage tables really seal the deal.
If you look at the Arabic example, already by the time you move to looking at words 1 edit distance away, there is hardly any discernable pattern, except for a tendency to not use similar words right after one another, and a slight tendency to use similar words at the same position in earlier lines (you see this in poetry—for example, if the ends of lines rhyme, you are going to see the line endings look like "ball," "fall," "tall," etc., or "cat," "sat," "hat." I don't know if the Quran was written a bit poetically, but it wouldn't surprise me. I bet if you had looked at words with edit distances of two in the Quran, the patterns would have disappeared entirely.
Whereas, with the VMS, the wider you cast the edit-distance net, the stronger the pattern becomes—like throwing a stone onto the page and seeing the similar words ripple out around it. Amazing.
The shaded percentage tables really seal the deal.
If you look at the Arabic example, already by the time you move to looking at words 1 edit distance away, there is hardly any discernable pattern, except for a tendency to not use similar words right after one another, and a slight tendency to use similar words at the same position in earlier lines (you see this in poetry—for example, if the ends of lines rhyme, you are going to see the line endings look like "ball," "fall," "tall," etc., or "cat," "sat," "hat." I don't know if the Quran was written a bit poetically, but it wouldn't surprise me. I bet if you had looked at words with edit distances of two in the Quran, the patterns would have disappeared entirely.
Whereas, with the VMS, the wider you cast the edit-distance net, the stronger the pattern becomes—like throwing a stone onto the page and seeing the similar words ripple out around it. Amazing.