Structural Signatures of the Voynich Manuscript

This deposit presents a structural and information‑theoretic study of the Voynich Manuscript (Beinecke MS 408), an undeciphered early‑fifteenth‑century codex. The work consolidates the established structural evidence for Voynichese into a single layered model — glyph, word, line, paragraph, and dialect — and reframes the manuscript's well‑known statistical signatures (first‑ and second‑order character entropy, the word‑length distribution, the Zipfian rank–frequency law, a two‑state hidden Markov decomposition, and line‑position effects) as a set of independent constraints. Treating them jointly converts a descriptive literature into a constraint‑satisfaction problem: which simple generative mechanisms can reproduce all of the signatures at once? Through controlled experiments on synthetic null models, the study shows that a finite‑state ("slot") word grammar combined with uniform word reuse can match any two of the three frequency‑related signatures but provably not all three, because a single reuse parameter is coupled to properties the manuscript holds at independent values. A length‑preserving reuse mechanism is then shown to resolve the conflict, reproducing the rank–frequency law and low conditional entropy while preserving the narrow, binomial word‑length distribution. The result is a falsifiable "decoupling principle" specifying what any complete generative account of the text must satisfy. All quantitative results are reported with their transcription and tokenisation choices. The generative findings concern structural logic on synthetic models and are not measurements of the manuscript itself; corpus validation against a real EVA transcription, partitioned by Currier system, is identified as the key next step. No claim is made regarding the meaning, language, or authorship of the manuscript. Cites: Amancio, D. R., Altmann, E. G., Rybski, D., Oliveira Jr., O. N., & Costa, L. da F. (2013). Probing the statistical properties of unknown texts: application to the Voynich Manuscript. PLoS ONE, 8(7), e67310. Bowern, C., & Lindemann, L. (2021). The linguistics of the Voynich Manuscript. Annual Review of Linguistics, 7, 285–308. Currier, P. H. (1976). Papers presented to a Voynich Manuscript seminar (in M. D'Imperio's collected materials). D'Imperio, M. E. (1978). The Voynich Manuscript: An Elegant Enigma. National Security Agency. Hodgins, G. (2009). Radiocarbon dating of the Voynich Manuscript parchment. University of Arizona AMS Laboratory. Landini, G. (2001). Evidence of linguistic structure in the Voynich Manuscript using spectral analysis. Cryptologia, 25(4), 275–295. Montemurro, M. A., & Zanette, D. H. (2013). Keywords and co-occurrence patterns in the Voynich Manuscript: an information-theoretic analysis. PLoS ONE, 8(6), e66344. Reddy, S., & Knight, K. (2011). What we know about the Voynich Manuscript. Proceedings of the 5th ACL-HLT Workshop on Language Technology for Cultural Heritage, Social Sciences, and Humanities, 78–86. Rugg, G. (2004). An elegant hoax? A possible solution to the Voynich Manuscript. Cryptologia, 28(1), 31–46. Rugg, G., & Taylor, G. (2016). Hoaxing statistical features of the Voynich Manuscript. Cryptologia, 41(3), 247–268. Stolfi, J. (2000). A grammar for Voynichese words (the crust-mantle-core paradigm). Online technical notes, IC-UNICAMP. Timm, T., & Schinner, A. (2020). A possible generating algorithm of the Voynich Manuscript. Cryptologia, 44(1), 1–19. Zandbergen, R. (n.d.). The Voynich Manuscript. voynich.nu. Zattera, M. (n.d.). A slot-based model of Voynichese word structure. v4j project notes. Keywrods: Voynich Manuscript, Beinecke MS 408, Voynichese, undeciphered script, historical cryptography, computational linguistics, quantitative linguistics, information theory, Shannon entropy, conditional entropy, Zipf's law, rank-frequency distribution, hidden Markov model, generative model, null model, finite-state grammar, word morphology, Currier languages, digital humanities, medieval manuscripts