Proposed architecture of the Voynich system

[Dunsel]

I ran some stats, and look at the results:
Semantic channel (o, a, t, k, p, f, d, l, r, n, m, y, s): stable across sections, preserved in captions, 73% of all glyphs
Elaboration channel (ch, sh, e, ee, q, i, ii, cXh): varies by line position, varies by A/B "language," largely absent from captions.

Thoughts?

If you've read my current post you know I'm working on a sheet based copy/mutate theory.  I found your idea curious so I fired up codex and using some of my data, I tested your findings across the astrologic section.  I tested this on pages (f67–f73) and can confirm the same general pattern. Labels/captions are dramatically more “core-heavy” and contain far less of the proposed elaboration layer than running text. The same effect also appears manuscript-wide.

Subset	Tokens	Semantic %	Elaboration %
Astrological labels	496	82.6%	14.9%
Astrological running text	2,130	73.7%	24.7%
All manuscript labels	575	83.0%	14.5%
All running text	32,982	70.4%	26.6%

Then I tested the proposed “core vs elaboration” idea by stripping the suggested elaboration glyphs/sequences (ch, sh, q, e, ee, i, ii) from the text and rerunning structural comparisons on Scribe 1.  The stripped forms actually collapsed MORE cleanly onto prior material. Exact/prior coverage increased substantially, edit distance 1 recovery improved, the number of source sheets needed dropped, and label vocabulary became noticeably closer to running text.

Test	Raw	Stripped
Scribe 1 exact/prior coverage	67.8%	80.5%
Scribe 1 exact+ED1 coverage	92.2%	96.5%
Sheets needed for 80% token coverage	3	2
Label/running-text vocabulary overlap	47.1%	56.9%

I don't think this proves semantics. But it does suggest there may really be a relatively stable underlying word-family backbone underneath a more variable elaboration layer.

And, that fits quite well with the constrained copy/mutate production process that I'm seeing. The “core” may simply represent inherited word-family anchors used during copying. While elaboration represents local mutation freedom.  In other words, the scribes are not preserving semantics, they are preserving recognizable production continuity.  That's a working theory mind you...

So, very interesting find,