Do f58r/f58v show Currier languages evolving?

[MarcoP]

A couple of years ago, I was inspired by Rene and Lisa's research to post CUVA bigram diagrams. You are not allowed to view links. Register or Login to view., based on the seven CUVA bigrams ED OL DY SO HO OR EO, is probably the most informative; this was also inspired by something Rene did, IIRC. You are not allowed to view links. Register or Login to view. shows diagrams for individual bigrams and details the color codes I used.

Indeed f58 is at a centre of a small cluster of folios at the centre of the PCA diagram: according to this graph, it can be seeen as being "mid-point". This small cluster is surrounded by a large CurrierA cloud at the bottom left, a large Currier B cloud at the top left and a (mostly) Pharmacese cloud on the right. This smaller cloud is made of 6 folios (58, 70, 51, 67, 68, 69 - the label for the red square 51 is obscured by the other overlapping folios) by three different scribes: s3 f58, s1 f51, s4 the other four folios.

Of course, PCA does not show the details. If one looks at the individual plots, these six folios are sometimes rather distant from each other.

These are the top ten closest folios based on PC1 and PC2 (Euclidean distance):

scribe f Section distance
4 70 ZodAstr 0.347
1 51 Botanic 0.370
4 67 ZodAstr 0.724
4 68 ZodAstr 0.783
4 69 ZodAstr 0.900
1 89 Recipes 2.086
1 44 Botanic 2.368
1 90 Botanic 2.676
1 53 Botanic 2.771
1 87 Botanic 3.007