Voynich Manuscript

Radiocarbon dating of the vellum places its production in the early fifteenth century, roughly 1404–1438. Ink and pigment analysis is consistent with that period and shows no sign of modern materials. The physical object is not a nineteenth-century forgery by its rediscoverer — that hypothesis is closed.

The text has structure. Word-length distributions, entropy measurements and the presence of Zipf-like frequency behaviour indicate something more organised than random glyph-stringing. Sections are conventionally grouped by illustration type: herbal, astronomical, balneological, pharmaceutical, recipes.

Proposed decipherments — proto-Romance, abbreviated Latin, Hebrew anagram, Turkic, Nahuatl, and others — typically demonstrate a reading of a handful of words or a single page, then fail to extend. The test is simple and rarely passed: a real solution should decode arbitrary new passages into coherent, contextually appropriate text without new ad-hoc rules per line.

The plants are the other problem. Most are not identifiable with real species, which undermines any decipherment that depends on matching labels to a known herbal tradition.

Three remain reasonable. A natural language in an unknown script or heavy abbreviation. A constructed or cipher language. Or elaborate glossolalic nonsense produced for sale — which would still have to explain the statistical structure, and that is harder than it sounds.