An attempt at extracting grammar from vord order statistics.

[davidd]

….so is replacing m for iin showing that it helps or worsens or doesn’t change voynich …. Lack of entropy ? Strict positional preference based on these glyphs? That’s what the challenge was right? Let’s see if we can mess with m as shorthand to see what happens. What happened? Not good at math here you  might need to um, paint a description? ??

In the tests it doesnt seem to change much about the score, which is a good sign that it may be true. If those are equivalent ways to write the same thing, the software should already be more likely to put those equivalent vords in the same group.