Given the structural characteristics I have measured over the past few months, a one-to-one substitution seems strong unlikely to me. The text is highly organized at the glyph level: there are distinct glyph families, and the number of strokes ( e,ee,eee, i, ii, iii) is part of the symbol’s identity rather than a free variation. Any proposed substitution must take this structure into account, yet this work does not address it at all.
The general point is simpler. A simple substitution preserves the statistical properties. Such a claim therefore implies that the plaintext exhibits the statistical profile of the Voynich text. This claim is never tested here—the key is applied to a few dozen lines and a series of labels, but never to contiguous text.
I cannot read Arabic, so I cannot assess the lexical aspect. However, the objection does not depend on Arabic: the issue is whether the proposed system is compatible with the structure of the ciphertext. The proof is missing and, by all accounts, will never be able to be provided.
The general point is simpler. A simple substitution preserves the statistical properties. Such a claim therefore implies that the plaintext exhibits the statistical profile of the Voynich text. This claim is never tested here—the key is applied to a few dozen lines and a series of labels, but never to contiguous text.
I cannot read Arabic, so I cannot assess the lexical aspect. However, the objection does not depend on Arabic: the issue is whether the proposed system is compatible with the structure of the ciphertext. The proof is missing and, by all accounts, will never be able to be provided.