Glyph Height Transition Entropy in the Voynich Manuscript

The Voynich Manuscript is famous for its unknown script and resistance to decipherment. While most research focuses on horizontal features such as word structure, glyph co-occurrence, and linguistic patterns, this study examines a purely vertical dimension: the transitions in glyph height. Using Takeshi’s EVA transcription of the entire manuscript, each glyph was classified into one of four height categories (A = ascender, M = middle, B = base, C = short/descender) based on visual inspection of the original pages. The sequence of height categories was then used to calculate Shannon entropy of the 16 possible height transitions (AA, AM, … CC) for each page. Analysis of 157,373 transitions across 225 pages reveals an average per-page transition entropy of approximately 3.49 bits — close to the theoretical maximum of 4.0 bits. More importantly, this high entropy value remains remarkably stable across all major sections of the manuscript (Botanical, Astronomical, Balneological, Cosmological, Pharmaceutical, and Recipe), despite their visual and structural differences. Notable findings include the near-total suppression of AA transitions (ascender followed by ascender) and the strong overrepresentation of AB transitions. These results indicate that ascenders appear almost exclusively as isolated peaks rather than in clusters. The combination of high entropy and exceptionally low variance across sections suggests the presence of consistent underlying rules governing height distribution — rules that go beyond what would be expected from a conventional alphabetic script or from pure randomness. This study provides quantitative evidence that the vertical dimension of the Voynich script is not incidental but follows a stable statistical pattern. The full dataset, including raw counts and per-page calculations, is provided in the accompanying spreadsheet for verification and further research. This upload contains: The full research paper (PDF) -The complete dataset with raw data and calculations (ODS)" To contact me: [e-mail]